Wide-temperature-range eccentric rotary table adjusting method and wide-temperature-range eccentric rotary table adjusting system

Abstract

The invention relates to the technical field of data processing, in particular to a wide-temperature-range eccentric rotary table adjusting method and system. The method comprises the following steps: starting an eccentric turntable in a high-temperature environment to obtain high-temperature reference matrix data; performing region division according to the high-temperature reference matrix data to obtain high-temperature reference partition data; performing eccentricity error calculation according to the high-temperature reference partition data to obtain eccentricity error data, and performing error change processing according to the eccentricity error data to obtain deviation error change data; and according to the deviation error change data, constructing an error characteristic graph to obtain high-temperature eccentric error characteristic data so as to carry out wide-temperature-range eccentric turntable adjustment auxiliary operation. According to the invention, the equipment can continuously and stably work in a high-temperature environment, so that the equipment has high-precision temperature compensation capability, the equipment adjustment process is optimized through data analysis, and the long-term operation stability and energy efficiency performance of the equipment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly relates to a calibration method and system for an eccentric turntable within a wide temperature range. Background Art

[0002] In high-temperature industrial environments, especially in fields such as oil drilling, logging, and mineral exploration, precise positioning control and orientation calibration are crucial for ensuring the safe operation and efficiency of equipment. In actual production, production equipment such as drilling tools, drill bits, and sensors operates in complex environments with extreme temperatures, high vibrations, and strong electromagnetic interference. Therefore, effectively calibrating the rotational accuracy and error compensation of these devices is an important technical means to ensure their long-term stable operation.

[0003] An eccentric turntable is a device commonly used for measuring and calibrating high-precision sensors, capable of simulating complex dynamic working conditions, especially for error calculation and compensation in high-temperature environments. It precisely controls the rotation and attitude of the turntable, uses built-in sensors (such as accelerometers, fluxgate sensors, and gyroscopes) to monitor the movement of the turntable in real time, and calibrates and adjusts the sensors through data processing. Most existing error calculation methods ignore the correlation between temperature changes and vibrations. Especially under the combined action of temperature gradients and vibration frequency changes, the error characteristics often exhibit non-linearity and time-variability, and traditional linear error correction methods cannot effectively compensate for these errors. Summary of the Invention

[0004] To solve the above technical problems, the present invention proposes a calibration method and system for an eccentric turntable within a wide temperature range to solve at least one of the above technical problems.

[0005] The present application provides a calibration method for an eccentric turntable within a wide temperature range, including the following steps: Step S1: Start the eccentric turntable in a high-temperature environment to obtain high-temperature reference matrix data; Step S2: Perform regional dissection based on the high-temperature reference matrix data to obtain high-temperature reference partition data; Step S3: Calculate the eccentric error based on the high-temperature reference partition data to obtain eccentric error data, and perform error change processing based on the eccentric error data to obtain deviation error change data; Step S4: Construct an error characteristic map based on the deviation error change data to obtain high-temperature eccentric error characteristic data for assisting in the calibration operation of the eccentric turntable within a wide temperature range.

[0006] In the present invention, by collecting detailed reference matrix data of the eccentric turntable and performing regional dissection under high-temperature environments, the performance changes of the device under high-temperature conditions can be captured more accurately. Under extreme working conditions such as high temperature and high vibration, the accuracy and stability of the eccentric turntable are guaranteed. Through eccentric error calculation and error change processing, the eccentric error of the device is dynamically adjusted, enabling the eccentric turntable to maintain high-precision output in different working environments and avoiding error accumulation and device performance degradation caused by high temperature. Analyzing the characteristics of error data can comprehensively evaluate the performance problems of the eccentric turntable. Compared with traditional methods, this multi-dimensional error analysis can more accurately identify the root causes of problems and optimize the calibration process. The construction of the error characteristic diagram makes the change trends and laws of the eccentric error more intuitive and understandable, helping to precisely adjust the turntable parameters during the calibration process, improving the calibration efficiency and quality. By performing calibration within a wide temperature range, the performance fluctuations of the device caused by temperature changes can be processed and compensated, enabling the eccentric turntable to work stably within a wider temperature range (such as from 85°C to 175°C), meeting the requirements for device performance in high-temperature environments in special industries such as oil drilling and mineral exploration.

[0007] Preferably, step S1 is specifically as follows: Step S11: Start the eccentric turntable in a high-temperature environment, and obtain initial high-temperature environment perception data through the sensors built in the eccentric turntable, where the sensors built in the eccentric turntable include a high-temperature accelerometer, a fluxgate, and a gyroscope; Step S12: Perform temperature stratification modeling on the initial high-temperature environment perception data to obtain a temperature level model; Step S13: Perform multi-modal fusion according to the temperature level model to obtain high-temperature fusion data; Step S14: Use an inertial measurement unit to record the rotation trajectory, extract the rotation characteristics under a high-temperature environment, and obtain rotation trajectory data; Step S15: Construct a data trajectory according to the high-temperature fusion data and the rotation trajectory data to obtain high-temperature reference matrix data.

[0008] In the present invention, through the combined use of high-temperature sensors such as high-temperature accelerometers, fluxgates, and gyroscopes, the dynamic information of the eccentric turntable in a high-temperature environment can be captured in real time, enabling stable operation in extreme environments such as high temperature and high vibration, providing accurate sensing data, and providing reliable basic data for the calibration process. The multi-modal data fuses the changes of various physical quantities, avoiding the limitations brought by single-sensor data, thereby improving the accuracy and integrity of environmental perception. The temperature stratification modeling can accurately distinguish the physical states at different temperatures, reasonably divide the temperature regions, avoid the influence of temperature gradients on the device performance, ensure that the temperature fluctuations during the data acquisition process are effectively considered, and lay a solid foundation for error analysis and calibration work. The establishment of the temperature hierarchy model helps to reduce the errors caused by temperature non-uniformity and enables the independent analysis of the physical properties in different temperature regions, thus more accurately describing the performance of the device in different working environments. The multi-modal fusion fuses the data of multiple sensors such as temperature, vibration, acceleration, and magnetic field to ensure that the change characteristics of the device can be comprehensively captured in a high-temperature environment, thereby improving the quality and consistency of the overall data, more accurately simulating the performance of the device in a complex environment, effectively enhancing the reliability of subsequent analysis and calibration, avoiding errors caused by differences between different data sources, and ensuring the consistency and integrity of the data. Precise rotation trajectory capture and dynamic analysis: By collecting dynamic data such as acceleration and angular velocity, the rotation behavior of the eccentric turntable can be accurately analyzed to obtain more detailed rotation characteristics. The rotation trajectory data obtained through the inertial measurement unit can effectively reveal the influence of the high-temperature environment on the rotation dynamics of the device, providing accurate motion parameters and supporting error analysis. By combining the high-temperature fusion data with the rotation trajectory data, high-temperature reference matrix data is generated, mapping the data of different sensors and different temperature regions to form an accurate and comprehensive high-temperature reference data model, ensuring that the most accurate device performance state can be obtained during the actual calibration process.

[0009] Preferably, step S2 is specifically as follows: Step S21: Calculate the temperature gradient and the vibration frequency according to the high-temperature reference matrix data to obtain the temperature gradient data and the vibration frequency data; Step S22: Divide the high-temperature reference matrix data according to the temperature gradient data to obtain the temperature gradient partition data; Step S23: Divide the high-temperature reference matrix data according to the vibration frequency data to obtain the vibration frequency partition data; Step S24: Conduct a change correlation analysis according to the temperature gradient data and the vibration frequency data to obtain the temperature-vibration correlation data; Step S25: Map the temperature gradient partition data and the vibration frequency partition data according to the temperature-vibration correlation data to obtain the temperature-vibration partition map data; Step S26: Perform error analysis and uncertainty analysis based on the temperature-vibration partition map data to obtain partition map error data and partition map uncertainty data respectively; Step S27: Perform partition optimization on the temperature-vibration partition map data according to the partition map error data and the partition map uncertainty data to obtain the first high-temperature reference partition data; Step S28: Perform multi-scale regional dissection based on the high-temperature reference matrix data to obtain the second high-temperature reference partition data.

[0010] In the present invention, the temperature gradient calculation and the vibration frequency calculation can accurately obtain the changes in the temperature gradient and the vibration frequency in a high-temperature environment. By partitioning the temperature gradient data and the vibration frequency data, accurate modeling and error calculation can be carried out for the environmental changes in different regions. Through the extraction of temperature-vibration correlation data, the relationship between temperature changes and vibration frequencies can be revealed. In a high-temperature environment, temperature and vibration often affect each other, and different vibration frequencies will result in different temperature distribution patterns. Error analysis and uncertainty analysis discover potential error sources, quantify and correct them, providing reliable data support for the calibration of the high-temperature eccentric turntable. Through multi-scale regional dissection, the calibration strategy can be dynamically adjusted according to different temperature and vibration ranges, improving the performance of the calibration method in a wide temperature range.

[0011] Preferably, step S21 is specifically as follows: Perform temperature difference calculation based on the high-temperature reference matrix data to obtain temperature difference data; Perform temperature gradient calculation based on the temperature difference data to obtain temperature gradient data; Perform Fourier transform on the high-temperature reference matrix data to obtain frequency-domain vibration data; Extract vibration frequencies based on the frequency-domain vibration data to obtain vibration frequency data.

[0012] In the present invention, the temperature difference calculation can accurately capture the minute changes in temperature in the high-temperature reference matrix data. Through differential processing, the irregularities and gradient changes in the temperature distribution can be revealed, providing a solid foundation for temperature gradient calculation, ensuring that the temperature changes in the high-temperature environment are comprehensively and accurately reflected, and providing key data for subsequent error correction. By processing the temperature difference data, the spatial distribution characteristics of temperature can be effectively revealed. Precise temperature gradient information can reflect the thermal behavior of equipment in a high-temperature environment, helping to further optimize the thermal management of the equipment and prevent performance degradation or equipment damage caused by uneven temperature. Under high-temperature conditions, the measurement of vibration frequency is particularly important because vibration is closely related to temperature changes. At high temperatures, temperature fluctuations affect the vibration characteristics of equipment. Fourier transform can convert complex time-domain signals into frequency-domain vibration data, revealing the vibration frequency characteristics of the equipment at different temperatures. By extracting features from the frequency-domain data, the natural frequency, resonance frequency, and other key vibration information of the equipment can be identified. This is crucial for judging the stability and health status of the equipment under high-temperature working conditions. Vibration frequency data helps to identify abnormal vibrations caused by temperature changes, thus providing early warnings of potential equipment failures or malfunctions. Through the joint analysis of temperature gradient and vibration frequency, fine-tuning of the dynamic behavior of equipment in a high-temperature environment can be achieved. Equipment under high temperatures will exhibit phenomena such as thermal expansion and physical deformation, resulting in changes in vibration frequency. The combined analysis of temperature difference and gradient data with vibration frequency data helps to predict the behavior of equipment under complex high-temperature conditions, detect potential performance degradation problems in advance, and thus provide data support for calibration.

[0013] Preferably, step S22 is specifically as follows: Construct a thermodynamic model of the brick well based on the temperature gradient data to obtain the thermodynamic model of the brick well; Extract the spatial variation characteristics of the temperature gradient based on the thermodynamic model of the brick well to obtain the spatial variation characteristic data of the temperature gradient; Set the temperature gradient threshold based on the spatial variation characteristic data of the temperature gradient to obtain the temperature gradient threshold data; Divide the high-temperature reference matrix data into temperature gradient regions through the temperature gradient threshold data to obtain the temperature gradient partition data.

[0014] In the present invention, by establishing a mathematical model between temperature gradient and thermodynamic characteristics, the thermal behavior of equipment in high temperature environment can be simulated and predicted more accurately. The model not only considers the temperature distribution, but also considers the influence of temperature change on equipment heat load, expansion and material properties, provides a theoretical basis for high temperature calibration of equipment, and helps to ensure the stability and reliability of equipment in extreme environments. The brick well thermodynamic model can capture the complex thermal phenomena such as heat conduction and convection of equipment under high temperature conditions, so that the subsequent calibration steps are more in line with the actual operation conditions and reduce the risk caused by thermal failure. Through in-depth analysis of temperature gradient data, the spatial variation characteristics of temperature gradient are extracted. Through this process, the hot spot areas of temperature change and their spatial distribution characteristics can be clearly identified. These characteristic data are helpful to find areas with large temperature gradient changes, which can effectively guide the thermal management and adjustment of equipment in high temperature environments and optimize equipment operating conditions. The extracted temperature gradient spatial variation characteristic data can provide decision support for temperature gradient threshold setting and regional division, ensure that the temperature response of different areas is accurately controlled, and avoid areas with excessive temperature changes affecting equipment performance. By determining the safety threshold of temperature gradient based on temperature gradient spatial variation characteristic data, a scientific basis is provided for subsequent regional division and calibration. Reasonable temperature gradient threshold settings can avoid excessive temperature differences during equipment operation, and prevent problems such as structural instability and material damage caused by overheating or excessive temperature differences. According to the working environment and material characteristics of the equipment, set appropriate temperature thresholds to optimize the equipment's operating strategy in high-temperature environments. By setting the temperature gradient threshold in advance, potential temperature unevenness can be warned and adjusted before the equipment is operated. Based on the set temperature gradient threshold data, different temperature gradient intervals can be accurately divided, and areas with more uniform temperature gradient distribution and sensitive areas with large temperature differences can be identified, which helps to accurately locate key areas in high-temperature environments, avoid long-term effects of high-temperature areas on equipment, and ensure that the equipment can operate stably in each area.

[0015] Preferably, step S23 is specifically: Perform wavelet transformation on the vibration frequency data to obtain first vibration frequency characteristic data; Calculating the frequency change rate of the vibration frequency data according to the temperature gradient spatial change characteristic data to obtain second vibration frequency characteristic data; Divide the frequency change interval according to the first vibration frequency characteristic data and the second vibration frequency characteristic data to obtain vibration frequency change interval data; The spatial region mapping is performed according to the vibration frequency variation interval data and the high temperature reference matrix data to obtain the vibration frequency partition data.

[0016] In the present invention, by processing the vibration frequency data through wavelet transform, the local features of frequency changes can be effectively captured, the instantaneous changes, mutations, and high-frequency noises hidden in the frequency can be identified, and the vibration frequency features caused by temperature changes, equipment stress, or external interference in a high-temperature environment can be accurately extracted. This step can transform the vibration data from the time domain to the frequency domain, facilitating the revelation of tiny changes inside the equipment and helping to identify potential vibration anomalies, such as eccentricity, mechanical looseness, etc., and predicting equipment failures in a timely manner. Analyze the vibration frequency data based on the data of the spatial variation characteristics of the temperature gradient, and quantify the dynamic correlation between temperature and vibration. The influence of temperature changes on the vibration frequency in a high-temperature environment affects the operating stability of the equipment and changes its working performance. By calculating the change rate of the vibration frequency, the direct influence of temperature changes on the vibration characteristics of the equipment can be better revealed, and the second vibration frequency characteristic data can be obtained, providing accurate data support for the correlation between the temperature gradient and the vibration characteristics. By combining the change of the temperature gradient with the change rate of the vibration frequency, a more comprehensive understanding of the dynamic behavior of the equipment under different temperature conditions can be achieved, especially how to optimize the design and adjustment in a high-temperature environment. Divide the vibration frequency data into intervals, so that different change ranges of the vibration frequency can be accurately represented. This process helps to identify high-frequency changes, low-frequency changes, and stable state regions, providing important references for subsequent tuning. By dividing according to the intervals of frequency changes, the abnormal frequency ranges that occur during the operation of the equipment can be identified. Especially when the temperature changes, the change trend of the frequency interval can help to judge whether the equipment is operating normally. The data after interval division will help to provide guidance in equipment maintenance, fault diagnosis, and preventive maintenance. Through refined interval division, more targeted optimization measures can be taken to improve the equipment stability and long-term operation efficiency. Combine the vibration frequency change interval with the high-temperature reference matrix data to provide spatial position data for the vibration characteristics in different regions under a high-temperature environment. The mapping helps to accurately identify different vibration frequency characteristics in the temperature change region and further analyze the operating states of different components of the equipment. Through mapping analysis, it can help to locate the regions with large frequency changes, providing a basis for precise tuning and optimization of the equipment.

[0017] Preferably, step S28 is specifically as follows: Perform non-linear image conversion according to the high-temperature reference matrix data to obtain high-temperature reference image data; Perform cross-scale space processing according to the high-temperature reference image data to obtain scale space data; Perform multi-scale blur processing on the scale space data through preset Gaussian scale space parameter data to obtain blurred scale image data; Perform scale gradient calculation according to the blurred scale image data to obtain scale gradient map data; Perform spatial feature extraction according to the scale gradient map data to obtain scale space feature data; Perform clustering calculation on the scale-space feature data to obtain scale-space region division data; Perform data mapping on the high-temperature reference matrix data according to the scale-space region division data to obtain the second high-temperature reference partition data.

[0018] In the present invention, cross-scale-space processing of the high-temperature reference image data can effectively reveal the variation characteristics of the data at different scales. Through cross-scale processing of the high-temperature reference matrix data, data structures and detail features at different levels can be discovered, and the data can be comprehensively analyzed from the macroscopic to the microscopic level. By performing non-linear transformation on the high-temperature reference matrix data, non-linear distortion in the data can be eliminated, making the image data under high-temperature environments more real and accurate. By introducing preset Gaussian scale-space parameter data for blurring processing, noise in the image can be processed and the key details of the data can be highlighted. Blurring processing helps to extract important features from the data at different scales. In high-temperature environments, it can effectively eliminate image noise caused by temperature or vibration. Performing scale gradient calculation on the blurred scale image data can reveal the subtle changes in the data and identify the variation trends of temperature, vibration, etc. at different scales, which helps to capture the performance of the device in different working environments. Especially in the face of complex high-temperature and vibration environments, it can better capture local abnormal changes. Through feature extraction, the influence of factors such as temperature and vibration on the device performance can be discovered at different levels, providing a more accurate basis for device calibration. Performing clustering processing on the scale-space feature data can automatically divide it into different regions according to the similarity of the data. Mapping the scale-space region division data to the high-temperature reference matrix data helps to achieve more accurate region recognition and calibration in high-temperature environments. By mapping the feature data to the specific working regions of the device, more specific and detailed adjustment suggestions can be provided, thereby optimizing the operation of the device at different times and in different environments.

[0019] Preferably, step S3 is specifically as follows: Step S31: Perform local eccentricity effect analysis based on the high-temperature reference partition data to obtain local eccentricity effect data; Step S32: Construct a temperature-vibration eccentricity error model based on the local eccentricity effect data to obtain the temperature-vibration eccentricity error model; Step S33: Perform local error calculation based on the high-temperature reference partition data and the temperature-vibration eccentricity error model to obtain eccentricity error data; Step S34: Perform time-series error analysis based on the eccentricity error data to obtain deviation error change data.

[0020] In the present invention, based on the high-temperature reference partition data, the eccentricity effect in different regions can be accurately identified, especially the local changes of the equipment in a high-temperature environment. The occurrence position of the eccentricity effect can be accurately located, avoiding misjudgment caused by the error of the overall eccentricity analysis. By comprehensively considering the influence of temperature and vibration on the eccentricity error, a comprehensive model is constructed to accurately describe the combined effect of the two on the eccentricity error. By combining the high-temperature reference partition data with the constructed temperature-vibration-eccentricity error model, the eccentricity error of each region can be accurately calculated, avoiding the averaging error in the overall eccentricity calculation. It can reveal the dynamic change process of the error. Especially under the action of high temperature and vibration, the eccentricity error has significant time-series change characteristics. It provides dynamic error data based on high-temperature conditions for the calibration process, making the calibration process of the equipment more accurate.

[0021] Preferably, step S4 is specifically as follows: Step S41: Perform error detrending according to the bias error change data to obtain detrended error data; Step S42: Perform bias error spectrum analysis on the detrended error data to obtain error spectrum data; Step S43: Perform error peak and zero-crossing point analysis on the error spectrum data to obtain error peak data; Step S44: Perform fuzzy clustering according to the error spectrum data and the error peak data to obtain error fusion feature data; Step S45: Perform feature map conversion according to the error fusion feature data to obtain high-temperature eccentricity error feature data for assisting the calibration operation of the wide-temperature-range eccentric turntable.

[0022] In the present invention, the long-term trend component in the bias error data can be removed to ensure that the analysis result only focuses on short-term fluctuations or non-linear changes. By transforming the error data into the frequency domain, the periodicity and frequency characteristics of the error changing with time can be revealed. By accurately calculating the key points (such as error peaks and zero-crossing points) in the error spectrum, it helps to identify the maximum change range and critical turning points of the error at different frequencies. Through the fuzzy clustering analysis of the error spectrum data and the peak data, the error characteristics can be classified into several groups, and then a clearer error pattern can be formed. The construction of the feature map makes the complex error data more intuitive and easy to understand, providing clear guidance for the adjustment in actual operation.

[0023] Preferably, the present application also provides a wide-temperature-range eccentric turntable calibration system for performing the wide-temperature-range eccentric turntable calibration method as described above. The wide-temperature-range eccentric turntable calibration system includes: A high-temperature reference matrix data acquisition module for starting the eccentric turntable in a high-temperature environment to obtain high-temperature reference matrix data; The high-temperature reference matrix region partitioning module is used to perform region partitioning based on the high-temperature reference matrix data to obtain high-temperature reference partition data; The bias error change calculation module is used to calculate the eccentricity error based on the high-temperature reference partition data to obtain eccentricity error data, and perform error change processing based on the eccentricity error data to obtain bias error change data; The high-temperature eccentricity error feature extraction module is used to construct an error feature map based on the bias error change data to obtain high-temperature eccentricity error feature data for assisting in the calibration of the eccentric turntable within a wide temperature range.

[0024] The beneficial effects of the present invention are as follows: By starting the eccentric turntable in a high-temperature environment and combining sensors to obtain initial high-temperature reference matrix data, accurate and reliable reference data is provided for the entire calibration process. The region partitioning of the high-temperature reference matrix data divides the data into regions according to specific rules, effectively decomposing complex factors such as variable temperature and vibration in the high-temperature environment into more operable regional data. The eccentricity error calculation and error change processing use the high-temperature reference partition data to accurately calculate the eccentricity error and perform error change processing on this basis. Through the construction of the error feature map, the bias error change data is converted into a visual error feature map, providing graphical auxiliary support for the calibration of the eccentric turntable within a wide temperature range. The error feature map can intuitively display the distribution characteristics of the eccentricity error under high-temperature conditions and reveal the working states of various regions of the device under different temperature and vibration conditions. Description of the Drawings

[0025] By reading the detailed description of the non-limiting embodiments with reference to the following drawings, other features, objectives, and advantages of the present application will become more obvious: Figure 1 Shows the step flow chart of a method for calibrating an eccentric turntable within a wide temperature range in an embodiment; Figure 2 Shows the step flow chart of a method for collecting high-temperature reference matrix data in an embodiment; Figure 3 Shows the step flow chart of a method for partitioning the high-temperature reference matrix region in an embodiment; Figure 4 Shows the step flow chart of a method for calculating the bias error change in an embodiment; Figure 5 Shows the step flow chart of a method for extracting high-temperature eccentricity error features in an embodiment. Detailed Embodiments

[0026] The technical method of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative work belong to the scope of protection of the present invention.

[0027] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0028] It should be understood that although terms such as "first", "second", etc. may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be called the second unit, and similarly the second unit may be called the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed associated items.

[0029] Please refer to Figures 1 to 5 , this application provides a calibration method for an eccentric turntable with a wide temperature range, including the following steps: Step S1: Start the eccentric turntable in a high-temperature environment to obtain high-temperature reference matrix data; Specifically, place the eccentric turntable in a high-temperature environment. Heat the turntable to the set high-temperature range through a heater and keep it stable. After starting the turntable, use an accurate angle sensor or rotary encoder to monitor the rotation of the turntable in real time, and record the actual eccentricity value at each rotation angle. This data is sampled multiple times to reduce accidental errors. During the implementation process, a series of temperature control points are set to ensure the stability of the high-temperature environment. The eccentricity data sampled each time is recorded as a matrix, where each matrix element represents the eccentricity value of the turntable at a specific temperature. The data structure of this matrix can be designed according to the number of sampling points and temperature intervals, and it is a two-dimensional matrix. The rows represent different temperature points, and the columns represent different rotation angles or times.

[0030] Step S2: Perform regional dissection based on the high-temperature reference matrix data to obtain high-temperature reference partition data; Specifically, after obtaining the high-temperature reference matrix data, the data is subjected to regional dissection to more precisely analyze the variation of the eccentricity error. Regional dissection decomposes the eccentricity error at high temperatures into variations in different regions, making the error calculation more accurate. The reference matrix is divided according to the temperature range, and a temperature partition standard is set (for example, every 10°C is an interval). For the data within each interval, the mean and standard deviation of the eccentricity error in this region are calculated. The matrix data within each temperature interval will be processed by standardization or normalization to remove the non-linear influence of temperature changes on the data. A set of reference partition data representing the variation of the eccentricity error within the temperature range of this interval will be obtained for each interval. During the dissection process, data fitting or regression analysis methods are used to accurately model the data to ensure that the data within each region can reflect the true error variation law, and high-temperature reference partition data is obtained, that is, a set of matrix data divided according to the temperature range.

[0031] Step S3: Calculate the eccentricity error based on the high-temperature reference partition data to obtain the eccentricity error data, and perform error variation processing based on the eccentricity error data to obtain the deviation error variation data; Specifically, according to the reference partition data, the eccentricity error within each temperature interval is calculated. The calculation method of the eccentricity error can be achieved by comparing the difference between the actually measured eccentricity value and the ideal eccentricity value. For example, calculate statistical quantities such as the maximum eccentricity error, root mean square error, or average error within each temperature interval. Perform variation processing on the eccentricity error. The purpose of the error variation processing is to identify the law of the error changing with temperature for compensation during calibration. Use data smoothing algorithms (such as weighted average method, moving average method, etc.) to process the error data of each temperature interval, remove noise or accidental fluctuations, and retain the actual influence of temperature changes on the eccentricity error, obtaining a set of deviation error variation data, that is, the trend change of the eccentricity error caused by temperature changes.

[0032] Step S4: Construct an error characteristic map based on the deviation error variation data to obtain high-temperature eccentricity error characteristic data for auxiliary operation of calibrating the wide-temperature-range eccentric turntable.

[0033] Specifically, an error feature map is constructed based on the deviation error change data obtained in step S3. The error feature map intuitively shows the law of the eccentricity error changing with temperature. A two-dimensional chart (such as a heat map, scatter plot, etc.) is used to represent the relationship between temperature and eccentricity error. The horizontal axis represents temperature, and the vertical axis represents the eccentricity error. The color depth or size of the data points can represent the magnitude of the error. Through the feature map, the change trend of the eccentricity error in different temperature ranges can be clearly seen. For more complex error features, methods such as piecewise fitting and interpolation are used to fit the data to generate a continuous error feature curve. Through this error feature map, it can help users judge the performance of the eccentricity error under different temperature conditions, thereby providing effective auxiliary data for the calibration of the wide-temperature-range eccentric turntable. During the calibration operation, the turntable can be precisely adjusted according to the error map to ensure that the eccentricity error within the entire temperature range remains within an acceptable range.

[0034] Preferably, step S1 is specifically as follows: Step S11: Start the eccentric turntable in a high-temperature environment, and obtain initial high-temperature environment perception data through the sensors built in the eccentric turntable. The sensors built in the eccentric turntable include a high-temperature accelerometer, a fluxgate, and a gyroscope; Specifically, start the eccentric turntable in a high-temperature environment to ensure that it is heated to a preset temperature range. Use a high-temperature furnace or heating device to heat the environment around the eccentric turntable to the required high-temperature state. The set temperature range can be from 85°C to 175°C. The sensors built in the eccentric turntable (including a high-temperature accelerometer, a fluxgate, and a gyroscope) will work simultaneously and start collecting environment perception data. The high-temperature accelerometer is used to detect the acceleration change of the turntable at different temperatures in real time and capture the vibration signal of the turntable during high-speed rotation. The data of the accelerometer will reflect the acceleration of the turntable in all directions, including linear acceleration and angular acceleration. The fluxgate sensor is used to measure the magnetic field change during the rotation of the turntable to help accurately capture the position change of the turntable. Through the subtle differences in the magnetic field change, the rotation state and eccentricity error of the turntable can be inferred. The gyroscope is used to detect the angular velocity and angle change of the turntable and can provide dynamic information during the rotation process to further help analyze the movement trajectory of the turntable in a high-temperature environment.

[0035] Step S12: Perform temperature stratification modeling on the initial high-temperature environment perception data to obtain a temperature level model; Specifically, the initially collected high-temperature environment perception data contains data at multiple temperature points, and each temperature point's data will have different error characteristics. To better analyze the impact of the high-temperature environment on the eccentric turntable, hierarchical modeling is performed according to the change of temperature data. Based on the collected temperature perception data, the temperature range is identified and the data is divided into different temperature levels. A temperature level standard is set (for example, every 10°C or every 5°C as a level) to segment the data. For the temperature data of each level, statistical modeling is carried out to fit the trend and law of the data, and data models for each temperature level are obtained through methods such as linear regression and quadratic fitting, reflecting the impact of temperature on the sensor output.

[0036] Step S13: Perform multimodal fusion according to the temperature level model to obtain high-temperature fusion data; Specifically, multiple sub-models are constructed according to temperature stratification. The models respectively reflect the change characteristics of sensor data at different temperatures. These different modal data are fused to obtain a comprehensive sensor response data set in a high-temperature environment. The model data at different temperature levels are fused using the weighted average method or other fusion algorithms. For the data of each temperature level, weighting is performed based on its reliability or weight in different temperature intervals. The weighted data are summed to obtain the fusion result. The fusion result at this time not only considers the data characteristics in each temperature interval but also combines the mutual influence between sensors to obtain a more accurate comprehensive data set. After multimodal fusion, it is also necessary to smooth the fusion data to remove potential noise or sharp change points, using low-pass filtering or mean filtering to ensure the smoothness of the fusion data in a high-temperature environment.

[0037] Step S14: Use an inertial measurement unit to record the rotation trajectory, extract the rotation characteristics in a high-temperature environment, and obtain rotation trajectory data; Specifically, an inertial measurement unit (IMU) is used to accurately record the rotation trajectory of the eccentric turntable. The inertial measurement unit combines the data of the accelerometer and the gyroscope and can accurately track the rotation path and eccentric state of the turntable. Start the eccentric turntable and record the acceleration and angular velocity data at each moment. The IMU will collect these data in real time and record the rotation state of the turntable in combination with the time stamp. By processing the IMU data, the rotation trajectory characteristics of the turntable are extracted, the kinematic characteristics such as the rotation angle, speed, and acceleration of the turntable are calculated, and its eccentricity is further analyzed. By integrating the angular velocity data of the gyroscope, the rotation angle of the turntable at different time points is obtained, and then combined with the acceleration data, the trajectory offset of the turntable is calculated.

[0038] Step S15: Construct a data trajectory according to the high-temperature fusion data and the rotation trajectory data to obtain high-temperature reference matrix data.

[0039] Specifically, the fusion data is combined with the rotation trajectory data to create a multi-dimensional data matrix. Each row in the matrix represents the fusion data of a temperature point, and each column represents different characteristics during the rotation process. The temperature level model, sensor data, and rotation characteristics are combined according to different dimensions of time or space. According to the relationship between the fusion data and the rotation trajectory, the multi-dimensional data matrix is filled. Each matrix element reflects the eccentricity data under a specific temperature and turntable angle, forming a high-temperature reference matrix, which serves as the basis for the calibration process. After the construction of the matrix is completed, the eccentricity error of the turntable under different temperature conditions can be accurately analyzed through this matrix, providing data support for the eccentricity calibration in a wide temperature range.

[0040] Preferably, step S2 is specifically as follows: Step S21: Calculate the temperature gradient and the vibration frequency according to the high-temperature reference matrix data to obtain the temperature gradient data and the vibration frequency data; Specifically, the temperature value of each data point is extracted from the matrix data, and the temperature difference (gradient) between adjacent data points is calculated. The temperature gradient refers to the rate of change of temperature between adjacent temperature points. By using the difference method, the gradient of each temperature point is calculated, and the forward difference or central difference method is used to calculate the temperature gradient. By performing frequency domain analysis on the accelerometer data, the vibration frequency of the turntable is extracted. The fast Fourier transform (FFT) is used to perform spectral analysis on the acceleration data to obtain a frequency distribution diagram. The vibration frequency is located at the peak position of the spectrogram. By identifying these peaks, the main vibration frequencies can be obtained. For example, if the amplitude at a certain frequency in the spectrogram is particularly large, it indicates that the turntable has strong vibration at that frequency.

[0041] Step S22: Divide the high-temperature reference matrix data according to the temperature gradient data to obtain the temperature gradient partition data; Specifically, according to the change of the temperature gradient, the threshold method or the clustering algorithm can be used for region division. A threshold range of the temperature gradient is set, and all regions with larger gradient values are divided into one group, and regions with smaller gradients are divided into another group. For example, a critical value of high and low gradients can be set, and regions with a temperature gradient higher than a certain threshold are marked as high-gradient regions, and those lower than the threshold are marked as low-gradient regions. By calibrating each temperature gradient value, the boundaries between different regions are identified to form the temperature gradient partition.

[0042] Step S23: Divide the high-temperature reference matrix data according to the vibration frequency data to obtain the vibration frequency partition data; Specifically, perform K-means clustering on the vibration frequency data with different numbers of clusters (K = 1, 2, 3,...). Calculate the silhouette coefficient for each cluster and obtain the average silhouette coefficient of all data points. The silhouette coefficient measures the tightness of a data point to points in its own cluster (i.e., the similarity of the data point to other points in the cluster, which is represented by calculating the average distance between the data point and all other data points in the same cluster) and its separation from data points in other clusters (i.e., the distinctiveness of the data point from other clusters, which is measured by the average distance between the data point and the data points in the other cluster that is closest to it). The value of the silhouette coefficient ranges from -1 to 1. The closer the value is to 1, the better the clustering effect (when the value of the silhouette coefficient is close to 1, it means that the data point is very similar to other points in the same cluster and the distance to the nearest cluster is very far, and the clustering effect is very good. When the value of the silhouette coefficient is close to 0, it means that the data point is on the boundary of the cluster, it has little difference from other points in the same cluster, and the distance to the nearest cluster is not far either, indicating that the clustering result is not clear. When the value of the silhouette coefficient is negative, it means that the data point is misclustered into the current cluster, and its distance to points in other clusters is closer than its distance to points in the current cluster). Select the number of clusters K corresponding to the maximum silhouette coefficient. For each data point, calculate its distance to all cluster centers, select the closest cluster center, and assign the data point to the corresponding cluster. For each cluster, calculate the mean of all data points in the cluster and update the cluster center to this mean. Iteratively calculate until the clustering result converges (i.e., the cluster centers no longer change). Through cluster analysis, divide the regions with the same or similar frequency characteristics into the same group. According to the frequency distribution in each region, the frequency response characteristics of each region can be further refined.

[0043] Step S24: Conduct a change correlation analysis based on the temperature gradient data and the vibration frequency data to obtain temperature-vibration correlation data; Specifically, analyze whether there is a significant linear or non-linear relationship between them by calculating the correlation coefficient between the temperature gradient and the vibration frequency. Calculate the Pearson correlation coefficient or the Spearman rank correlation coefficient to evaluate the relationship between the temperature gradient and the vibration frequency. For example, a higher temperature gradient will cause a change in the vibration frequency, so this change needs to be quantified. To more accurately capture their correlation, use multiple regression analysis or principal component analysis (PCA) to synthesize the data of the temperature gradient and the vibration frequency into a combined feature.

[0044] Step S25: Map the temperature gradient partition data and the vibration frequency partition data according to the temperature-vibration correlation data to obtain temperature-vibration partition map data; Specifically, according to the characteristics of the temperature gradient partition and the vibration frequency partition, as well as their correlation, a two-dimensional partition map is constructed. The horizontal axis of the map represents the temperature gradient, and the vertical axis represents the vibration frequency. Different regions are distinguished by colors or markings. For example, regions with a higher temperature gradient and a larger vibration frequency correspond to certain specific eccentricity error characteristics. The partition map is constructed by mapping the correlation data to generate a temperature-vibration partition map. Each region in the map represents a different combination of temperature gradient and vibration frequency.

[0045] Step S26: Conduct error analysis and uncertainty analysis based on the temperature-vibration partition map data to obtain partition map error data and partition map uncertainty data respectively; Specifically, for each partition, calculate the error magnitudes caused by the temperature gradient and the vibration frequency. By comparing the deviation between the actual measurement data and the theoretical model, calculate the average value and standard deviation of the error. Based on the error analysis, further analyze the measurement uncertainty of different partitions. Use methods such as Monte Carlo simulation to simulate the error propagation under different temperature and vibration conditions, so as to evaluate the overall uncertainty of the system.

[0046] For each partition, calculate the errors caused by the temperature gradient and the vibration frequency. The error refers to the deviation between the actual measured value and the theoretically calculated value, that is, in a high-temperature environment, the difference between the temperature value measured by the sensor and the theoretically calculated temperature value. Under different vibration frequencies, the difference between the measured vibration frequency value and the theoretically predicted value. Select a set of measurement data under the same temperature and vibration conditions and record the actual measured values. Calculate the theoretically predicted values through the temperature gradient model and the vibration frequency calculation model. Calculate the error, that is, the actual measured value minus the theoretically predicted value, to obtain the deviation data. To analyze the concentration degree of the error in different partitions, calculate the mean value of the error. The mean value reflects the overall deviation trend of the temperature and vibration frequency within this partition. Calculate the sum of the errors of all measurement points within this partition. Calculate the number of measurement points. Divide the sum of the errors by the number of measurement points to obtain the average value of the error. The standard deviation is used to measure the fluctuation range of the error, indicating the dispersion degree of the measurement error within this partition, that is, whether the error is relatively concentrated or there are large variations. Calculate the deviation between the error of each measurement point and the mean value of the error. Calculate the sum of the squares of the deviations of all measurement points. Divide the sum of the squares of the deviations by the number of measurement points, and then take the square root to obtain the standard deviation of the error.

[0047] Error propagation analysis is used to calculate the impact of different error sources on the measurement results. Error propagation means that errors do not affect the measured values independently, but will be superimposed on each other to form measurement errors. Calculate the impact degrees of temperature gradient error and vibration frequency error on the measurement results, and determine their contribution ratios. Calculate the impact of temperature gradient error on the measured value. Calculate the impact of vibration frequency error on the measured value. Combine the impacts of different errors to calculate the comprehensive error of the measurement system. Since the propagation of errors is non-linear, the Monte Carlo simulation method is used to estimate the uncertainty of the system. Monte Carlo simulation is a method for approximately calculating the errors of complex systems through random sampling. Set the distribution ranges of temperature gradient error and vibration frequency error (such as normal distribution). Conduct a large number (e.g., 1000 - 10000 times) of random samplings to simulate the error propagation process under different temperature and vibration conditions. Statistically analyze all simulation results and calculate the uncertainty range of the measurement system, including the measurement mean and standard deviation. The uncertainty data of the partition map includes the error propagation analysis results of temperature gradient and vibration frequency, the error distribution data obtained through Monte Carlo simulation, and the calculated uncertainty of the measured value (mean, standard deviation).

[0048] Step S27: Optimize the temperature-vibration partition map data according to the partition map error data and the partition map uncertainty data to obtain the first high-temperature reference partition data; Specifically, use the error data and uncertainty data to optimize the temperature-vibration partition map, reduce errors and uncertainties, and obtain more accurate reference partition data. By adjusting the thresholds of temperature gradient and vibration frequency, optimize the boundaries of each partition to minimize errors and reduce the impact of uncertainties.

[0049] Step S28: Perform multi-scale regional dissection on the high-temperature reference matrix data to obtain the second high-temperature reference partition data.

[0050] Specifically, use analysis methods at different scales to refine the matrix data. After rough partitioning at a large scale, further divide each partition at a small scale. By reducing the analysis scale, more subtle error distributions can be identified.

[0051] Specifically, a fixed grid division method is adopted for the entire high-temperature reference matrix, for example, it is divided into an initial grid area of 10×10. Each grid contains multiple measurement points, and the average temperature, temperature gradient, vibration frequency, and their change rates within this area are calculated to form a large-scale temperature-vibration feature vector. Subsequently, the K-means clustering or gradient-based region growing algorithm is used to cluster each grid area according to the similarity of temperature-vibration characteristics, obtaining a set of preliminary large-scale partition data. For each large-scale area, the internal temperature change range ΔT and the vibration error change range ΔV are calculated. When ΔT is greater than the preset temperature change threshold (such as 5°C), or ΔV is greater than the preset vibration error threshold (such as 0.1 mm), it is marked as an area to be subdivided. The above-mentioned large areas marked as to be subdivided are further refined into smaller sub-areas, for example, from a 10×10 grid to a 5×5 grid, or even a 2×2 grid. The local temperature gradient and vibration error gradient are calculated within the sub-areas. If the gradient change still exceeds the set threshold, recursive finer-level region division is performed until the stability determination condition is reached or the minimum division scale is reached. The division results at each scale are combined to generate the second high-temperature reference partition data containing multiple levels and different granularities.

[0052] Preferably, step S21 is specifically as follows: Perform temperature difference calculation based on the high-temperature reference matrix data to obtain temperature difference data; Specifically, obtain the temperature value at each position in the high-temperature reference matrix. Each element of the matrix represents the temperature data under specific time and temperature conditions. Temperature difference refers to the temperature change amount between adjacent points. By performing differences on each pair of adjacent data points, the change rate of temperature is obtained. The difference method can adopt forward difference, backward difference, or central difference.

[0053] Perform temperature gradient calculation based on the temperature difference data to obtain temperature gradient data; Specifically, the temperature gradient is the speed of temperature change at each position and is calculated through the temperature difference data. In a two-dimensional space, the gradient calculation processes the horizontal and vertical differences separately. By calculating the gradient at each point, a data set containing temperature gradient information can be obtained.

[0054] Perform Fourier transform based on the high-temperature reference matrix data to obtain vibration data in the frequency domain; Specifically, the accelerometer data is extracted from the high-temperature reference matrix, and this data contains the acceleration information of the turntable at different time points. The acceleration data is the vibration information at different frequencies and includes the dynamic response of the turntable under different temperature conditions. The Fourier transform is used to convert the acceleration data from the time domain to the frequency domain. The goal of the Fourier transform is to decompose the acceleration signal in the time domain into individual frequency components. Through the Fourier transform, the amplitude and phase information of the turntable vibration signal at each frequency can be obtained, thereby identifying the main vibration frequencies of the turntable.

[0055] Vibration frequency extraction is performed based on the frequency-domain vibration data to obtain vibration frequency data.

[0056] Specifically, a fast Fourier transform (FFT) is performed on the time-domain vibration signal collected in the high-temperature environment to obtain frequency-domain vibration data. There are multiple significant peaks in the frequency-domain data, and these peaks correspond to the resonance response characteristics of the eccentric turntable at specific frequencies. A local maximum detection algorithm (such as judging based on the extreme points of the first-order difference or scanning for peaks in a sliding window) is used to identify the positions of all the main peaks in the spectrum. To avoid interference from false responses, an energy threshold and a minimum peak distance can be set to filter out the influence of low-amplitude or high-frequency noise. Combining the spectrum and temperature information, the current temperature environment parameters are marked for each set of frequency peak data, and a temperature-frequency coupling index model is established. The extracted vibration frequency data set not only contains the main vibration frequency points under different temperature conditions, which is the vibration frequency data.

[0057] Specifically, further, a time-domain spectral transformation and wavelet packet energy analysis are performed on the vibration signal to obtain the main frequency peak in the frequency domain and the center frequency of the main mode of the wavelet packet time-frequency respectively; the modal confidence index of the candidate frequency point is calculated. , the The calculation formula is: , where is the modal confidence index of the candidate frequency point, is the peak amplitude of this frequency point in the frequency domain, is the standard deviation of the background noise frequency domain segment, is the frequency half-power bandwidth, is the center frequency of this frequency, is the wavelet packet band energy corresponding to this frequency, is the total wavelet packet energy of the entire signal. When the value is greater than the preset confidence threshold (such as ), and it is in the local maximum region in both the spectrum and the wavelet packet mode, then this frequency point can be determined as the effective main vibration frequency; further, the frequency results extracted from the time domain and the wavelet packet frequency domain are cross-validated, and only the consensus frequency points that are verified to pass in both bimodal modes are retained to obtain the vibration frequency data.

[0058] Preferably, step S22 is specifically as follows: Construct a thermodynamic model of the brick well based on the temperature gradient data to obtain the thermodynamic model of the brick well; Specifically, according to the temperature gradient data, the heat conduction process under different temperature conditions satisfies certain basic thermodynamic laws. For example, according to Fourier's law of heat conduction, the rate of heat transfer in each direction is proportional to the temperature gradient. According to the basic principle of the thermodynamics of the brick well, the relationship between the heat flux density and the temperature gradient is expressed by an equation. , where is the heat flux density, is the thermal conductivity of the material, is the temperature gradient. The heat flux density and the temperature gradient are directly related. Based on the thermodynamic model of the brick well, using the known temperature gradient data and thermal conductivity information, a simulated temperature field is established, and numerical solutions are carried out by the finite difference method or the finite element method to obtain the distribution of the temperature field, so as to establish a thermodynamic model of the brick well describing the spatial distribution of temperature.

[0059] Extract the spatial variation characteristics of the temperature gradient based on the thermodynamic model of the brick well to obtain the spatial variation characteristic data of the temperature gradient; Specifically, by performing a spatial analysis on the output results (i.e., temperature distribution) of the thermodynamic model of the brick well, the variation characteristics of the temperature gradient are extracted, and a local analysis of the temperature gradient data is carried out in space to evaluate the temperature gradient changes in different regions. For example, calculate the temperature change rate of each region, or use the gradient vector to describe the direction and magnitude of the temperature change. After extracting the variation characteristics of the local temperature gradient, methods such as cluster analysis and principal component analysis (PCA) are used to further analyze the spatial variation of the temperature gradient. Through the process of extracting the spatial variation characteristics, the spatial variation characteristic data of the temperature gradient are obtained.

[0060] Set the temperature gradient threshold based on the spatial variation characteristic data of the temperature gradient to obtain the temperature gradient threshold data; Specifically, threshold setting is a key step in region division. An empirical-based method or a statistical method is used to set the threshold. For example, according to the distribution of the temperature gradient, upper and lower limits are selected to delimit different regions. The temperature gradient values of all measurement points are statistically analyzed to construct a normal distribution, and the threshold is set based on the mean and standard deviation. Set the threshold for the high-temperature gradient region. If the temperature gradient is greater than the mean plus twice the standard deviation, then this region belongs to the high-temperature gradient region. Set the threshold for the low-temperature gradient region. If the temperature gradient is less than the mean minus one standard deviation, then this region belongs to the low-temperature gradient region. The temperature gradient data are divided into a high-gradient region (rapid change region), a low-gradient region (stable region), and an intermediate region (transition region).

[0061] Specifically, if the temperature gradient changes very sharply, a higher threshold is set to divide these regions into an independent region. According to the spatial variation characteristics of the temperature gradient, the temperature gradient ranges of each region are analyzed, clustering calculations are performed, and classification is carried out according to three clusters. For example, by observing the temperature gradient distribution map, it is determined which regions have a large change in temperature gradient (for example, when exceeding a certain critical value) and are set as "high-gradient regions"; while regions with a small change in temperature gradient can be set as "low-gradient regions". After setting the temperature gradient threshold, temperature gradient threshold data is obtained.

[0062] The high-temperature reference matrix data is divided into temperature gradient regions through the temperature gradient threshold data to obtain temperature gradient partition data.

[0063] Specifically, using the temperature gradient threshold data, regions are divided by comparing whether the temperature gradient of each region exceeds the set threshold. For example, if the temperature gradient of a certain region is greater than the set threshold, that region is marked as a high-gradient region; otherwise, it is marked as a low-gradient region. The temperature gradient characteristics of each region are automatically calibrated according to the previously set threshold. For example, the number of the high-gradient region is set to 1, and the number of the low-gradient region is set to 2. By checking and classifying each point in the matrix, the identification data of each region is obtained. According to the result of the region division, a temperature gradient partition map data is generated. The partition map shows the temperature gradient variation characteristics of each region in the high-temperature reference matrix.

[0064] Preferably, step S23 is specifically: Wavelet transform is performed on the vibration frequency data to obtain first vibration frequency characteristic data; Specifically, a wavelet basis function is selected according to the characteristics of the vibration frequency data, such as the Morlet wavelet or the Daubechies wavelet. The wavelet basis function can perform a good transformation between the time domain and the frequency domain and is particularly suitable for analyzing signals with frequency changing over time. Applying wavelet transform to the vibration frequency data gives a time-frequency distribution. The signal of the vibration frequency is transformed from the time domain to the wavelet domain to obtain a series of wavelet coefficients, and each wavelet coefficient represents the local frequency component of the signal within a certain time window. For each wavelet coefficient, its corresponding frequency component is calculated, and the frequency changes at different time points are calibrated. This helps to capture the instantaneous changes of the frequency in different time periods. From the coefficients obtained by the wavelet transform, an interested frequency interval is selected, such as the part with large frequency fluctuations, and the first vibration frequency characteristic data is extracted.

[0065] The frequency change rate of the vibration frequency data is calculated according to the temperature gradient spatial variation characteristic data to obtain second vibration frequency characteristic data; Specifically, the frequency change rate reflects the dynamic change of the vibration frequency during the temperature change process. By comparing the vibration frequency data at different time points or spatial regions, the change rate of the vibration frequency is calculated.

[0066] Specifically, when calculating the frequency change rate of the vibration frequency, the vibration frequency data at different time points and spatial regions are compared to measure the influence of temperature change on the vibration frequency. In the time dimension, the vibration frequency changes at the same position at different time points are compared, and the corresponding temperature change amount is calculated. By calculating the frequency change amount between two time points divided by the temperature change amount, the frequency change rate in the time dimension at this point can be obtained. Similarly, in the spatial dimension, the vibration frequency change amounts at adjacent spatial positions are compared, and the frequency change rate in the spatial dimension is calculated in combination with the temperature change (for weight calculation) between these positions. In order to obtain the frequency change trend of the entire region, the frequency change rate is calculated at multiple measurement points, and the data within the entire partition is statistically analyzed. The average value of the frequency change rates of all measurement points is calculated to obtain the average change trend of the entire partition. The maximum and minimum change rates are extracted to identify the regions with the most significant frequency change and the most stable regions. After the calculation is completed, the second vibration frequency characteristic data is obtained, which describes the dynamic change of the vibration frequency under different spatial variations of the temperature gradient, including the average change rate, change range, and trend characteristics of each partition.

[0067] Based on the first vibration frequency characteristic data and the second vibration frequency characteristic data, the frequency change interval is divided to obtain the vibration frequency change interval data; Specifically, by analyzing the first and second vibration frequency characteristic data, it is determined which frequency ranges have more significant changes. The regions with larger frequency changes indicate that the vibration response of the turntable is stronger within this temperature range, while the regions with smaller frequency changes indicate weaker vibration responses. A threshold range is set, and based on this threshold, the frequency change intervals are divided. For example, the part where the frequency change rate exceeds a certain threshold can be regarded as the high-frequency change interval, and the part below the threshold is the low-frequency change interval. According to the set threshold, the first and second vibration frequency characteristic data are divided into intervals. Each frequency change interval corresponds to a specific temperature condition and vibration response mode. By classifying the frequency changes, the frequencies with large changes are divided into the high-frequency change interval, and the frequencies with small changes are divided into the low-frequency change interval. The division criteria of the intervals are obtained through statistical analysis. For example, statistical quantities such as the standard deviation and mean of the data are used to set the threshold. Through these divisions, the vibration frequency change interval data is obtained.

[0068] Based on the vibration frequency change interval data and the high-temperature reference matrix data, a spatial region mapping is performed to obtain the vibration frequency partition data.

[0069] Specifically, based on the vibration frequency change interval data and combined with the spatial coordinates in the high-temperature reference matrix, the vibration frequency of each region is partitioned and mapped to a specific spatial region. Through this mapping, a vibration frequency category (such as a high-frequency region or a low-frequency region) is assigned to each region. The mapping rule is based on the previously divided frequency change intervals. For example, if the frequency change rate of a certain region exceeds the set threshold, this region will be calibrated as a high-frequency region; if the frequency change rate is low, it will be calibrated as a low-frequency region. After the spatial region mapping, the obtained vibration frequency partition data reflects the vibration characteristics of different regions in the high-temperature environment.

[0070] Preferably, step S28 is specifically as follows: Perform non-linear image conversion based on the high-temperature reference matrix data to obtain high-temperature reference image data; Specifically, through non-linear conversion, certain features in the high-temperature data can be enhanced, such as regions that are more sensitive to temperature changes. The conversion is carried out by introducing non-linear functions such as logarithms, power laws, or exponents. Set the conversion function (such as a logarithmic function) to process each data point in the high-temperature reference matrix. For example, if a certain temperature value in the original matrix is , then after non-linear image conversion, the value of the corresponding point in the new image can be obtained through the following relationship: , or use the power-law relationship: , where is an adjustment parameter term. After non-linear conversion, the obtained high-temperature reference image data has more details and features.

[0071] Perform cross-scale spatial processing on the high-temperature reference image data to obtain scale-space data; Specifically, the scale space refers to the process of analyzing different levels of detail by changing the scale of the image (by smoothing or blurring the image). For the high-temperature reference image, different scale representations of the image are obtained by changing the image resolution or by applying filters of different sizes. Use the high-temperature reference image data to process the image at different scales. Apply some smoothing filters (such as Gaussian filters) to blur the image to different degrees to obtain images at different scales. As the scale increases, the details in the image will be gradually blurred, and larger structures and trends are captured. After cross-scale spatial processing, the obtained scale-space data contains representations of the image at different scales, providing multi-level image features.

[0072] Specifically, through heat-gradient-guided scale-space processing, different regions set different Gaussian kernel widths according to the temperature change rate. A fine scale is adopted for regions with intense thermal fluctuations, and a large scale is used for blurring in thermally stable regions, enhancing the physical adaptability of the image scale-space representation. For each scale image, its temperature edge map (gradient magnitude) and regional stability index (local variance) are calculated to form a scale feature tensor. The obtained scale-space data is used as input for multi-scale clustering region dissection in the steps to achieve hierarchical mapping of the high-temperature matrix and extraction of error features.

[0073] By constructing the scale space of the image, temperature and structural features at different levels are extracted. Regarding the high-temperature reference image data as a two-dimensional matrix, it is smoothed at multiple levels using a Gaussian filter, with different levels corresponding to different Gaussian kernel radii (e.g., = 1, 2, 4, 8), obtaining a multi-scale image sequence. Each image corresponds to the structural representation at a scale. As the scale increases, the image details are gradually erased, and what remains are regional temperature patterns or structural contours. Using operators such as Sobel and Laplacian on the high-temperature reference image, the gradient magnitude map of each pixel is obtained; the mapping relationship between the temperature gradient and the filtering scale is set as: ; where is the filtering scale data, is the minimum filtering scale, is the maximum filtering scale, is the temperature gradient at this point, is the maximum value of the temperature gradient; the image is divided into blocks (e.g., 10×10), and each block selects different values according to its average (i.e., ) for Gaussian filtering; or a spatially variable filter is used with the support of a GPU; the generated multi-scale image data shows high-resolution details in regions with intense thermal motion and a compression trend in thermally stable regions.

[0074] Multi-scale blurring processing is performed on the scale-space data according to the preset Gaussian scale-space parameter data to obtain blurred scale image data; Specifically, the Gaussian scale space analyzes features at different scales by continuously blurring the image. Using the preset Gaussian scale-space parameter data (such as the standard deviation of the Gaussian function) to control the intensity of image blurring. Applying Gaussian blurring to the scale-space data and performing multi-scale processing with different standard deviation values of the Gaussian function. After Gaussian blurring processing, the obtained blurred scale image data has a good smoothing effect, eliminating high-frequency noise while retaining the main features of the image.

[0075] Calculate the scale gradient according to the blurred scale image data to obtain the scale gradient map data; Specifically, the scale gradient reflects the change rate of the image at different scales. By calculating the local differences of the image at multiple scales, important features in the image (such as edges, textures, etc.) can be revealed. The scale gradient is obtained by calculating the derivative of the blurred scale image in space.

[0076] Extract spatial features according to the scale gradient map data to obtain scale space feature data; Specifically, by analyzing the local changes in the scale gradient map, some key spatial features are extracted. For example, regions with large gradient changes are extracted as edge features, or regions with small gradients are extracted as flat regions. Spatial features include edges, corners, textures, etc. Edge detection, corner detection and other methods are used to extract the spatial features of interest from the scale gradient map. For example, the Canny edge detection method is used to extract edge features, or the Harris corner detection method is used to extract corner features.

[0077] Perform clustering calculation on the scale space feature data to obtain scale space region division data; Specifically, the K-means clustering method is used to divide the scale space feature data into different clusters. Each cluster corresponds to a region in the image with similar scale space features. By calculating the distance from each data point to the cluster centers, similar regions are grouped into one category. Through clustering calculation, the scale space region division data is obtained, and the image is divided into multiple regions with similar features.

[0078] Perform data mapping on the high-temperature reference matrix data according to the scale space region division data to obtain the second high-temperature reference partition data.

[0079] Specifically, by applying the clustering results to each data point in the high-temperature reference matrix, the scale space features of each region are mapped. The scale space category corresponding to each region can be assigned to the corresponding data point in the high-temperature reference matrix. Through this mapping, the obtained second high-temperature reference partition data reflects the characteristics of regions at different scales, enabling the high-temperature data to be analyzed at a higher level of resolution.

[0080] Preferably, step S3 is specifically as follows: Step S31: Perform local eccentricity effect analysis according to the high-temperature reference partition data to obtain local eccentricity effect data; Specifically, the local eccentricity effect refers to the phenomenon that the rotation or position deviation of the turntable is caused by the influence of temperature change or vibration on the turntable. The local effect is manifested in certain specific regions, and the eccentricity effects in different regions are different. According to the high-temperature reference partition data, analyze the temperature change situation in each region, and combine with the vibration frequency data extracted previously to judge the eccentricity effect in the local region. In the high-temperature reference partition, the temperature gradient and vibration frequency in each region are different, which will directly affect the eccentricity effect in that region. For each partition region, calculate the temperature gradient within that region. A larger temperature gradient means a stronger local eccentricity effect. According to the vibration frequency data, evaluate the relationship between the vibration frequency and the temperature gradient within the region. There is a certain correlation between the change in vibration frequency and the change in temperature gradient, which reflects the dynamic eccentricity effect in the local region. Combining the above analysis, for each partition region, calculate its eccentricity effect intensity, , is the intensity of the local eccentricity effect, is the temperature gradient weight factor, is the temperature gradient, is the vibration frequency change weight factor, is the change in vibration frequency, is the angular deviation weight factor, is the angular deviation, =0.03 mm / (°C / m), =0.02 mm / Hz, =1.5 mm / rad. Or, the eccentricity effect is represented by quantization indexes (such as rotation angle deviation, position deviation) to obtain the local eccentricity effect data.

[0081] Step S32: Construct a temperature-vibration eccentricity error model based on the local eccentricity effect data to obtain the temperature-vibration eccentricity error model; Specifically, according to the local eccentricity effect data, identify the contributions of temperature change and vibration frequency to the eccentricity effect. The expansion caused by the temperature gradient and the dynamic influence of the vibration frequency need to be considered in combination. The eccentricity error is related to the changes in the temperature gradient and the vibration frequency, and the influences of these two on the eccentricity error are superimposed. Select methods such as linear regression, weighted average, etc., with the temperature gradient and vibration frequency as inputs and the eccentricity error as the output to obtain the eccentricity error model. Or, use the temperature gradient, vibration frequency, and second-order coupling term as input features; use the measured eccentricity error (such as measured during turntable calibration) as the output; , where is the predicted eccentricity error, is the temperature gradient coefficient, is the temperature gradient, is the vibration frequency coefficient, is the change in vibration frequency, is the cross-coupling coefficient, is the offset constant term, and the model parameters are obtained by fitting historical training data as = 0.04 mm / (°C / m), = 0.025 mm / Hz, = 0.002 mm / (°C·Hz), = 0.12 mm. The model is trained and fitted with experimental or simulation data, and indicators such as mean squared error (MSE) and maximum residual are used to evaluate the model performance. When the deviation between the predicted result and the measured error is controlled within ±0.01 radian or 0.5 mm, it is considered that the model accuracy meets the engineering calibration requirements. By comparing the actual measured eccentricity error data with the model-predicted data, the model parameters are adjusted to obtain a more accurate temperature-vibration eccentricity error model.

[0082] Step S33: Calculate the local error according to the high-temperature reference partition data and the temperature-vibration eccentricity error model to obtain the eccentricity error data; Specifically, the temperature gradient and vibration frequency data of each region are extracted from the high-temperature reference partition data. According to the temperature-vibration eccentricity error model constructed in, the temperature gradient and vibration frequency of each region are used as inputs and substituted into the model for calculation. The eccentricity error data of each region are obtained. If there is new high-temperature reference partition data or environmental changes, the model can be dynamically updated to obtain more accurate eccentricity error data.

[0083] Step S34: Perform time-series error analysis on the eccentricity error data to obtain the deviation error change data.

[0084] Specifically, according to the eccentricity error data at each time point, its change rate is calculated. By analyzing the error change rate, the long-term change trend of the error can be identified. If the error shows a linear growth or decay trend, the growth or correction process of the eccentricity error can be inferred. If the error change has no pattern, it indicates that there are unstable factors in the system. By establishing a time-series model of the error (such as weighted moving average, exponential smoothing, etc.), the change trend of the error is further smoothed and predicted to obtain more stable error change data. After time-series analysis, the obtained deviation error change data will help identify the change pattern of the error in the time dimension.

[0085] Preferably, step S4 is specifically: Step S41: De-trend the error according to the deviation error change data to obtain the de-trended error data; Specifically, detrending refers to removing the trend component of a time series from the data to make the data more stable. A model (such as linear regression or polynomial regression) is fitted to capture the long-term trend in the data. A regression fit is performed on the data with biased error variation to obtain the trend part. By subtracting the trend part from the original data, detrended error data is obtained. The detrended error data will remove the influence of the long-term trend and only retain short-term fluctuations and outliers.

[0086] Step S42: Perform a biased error spectrum analysis on the detrended error data to obtain error spectrum data; Specifically, perform a Fourier transform on the detrended error data to obtain frequency domain data. Through this transform, the error signal in the time domain is converted into the amplitude representation of each frequency component. Extract the frequency range and amplitude information of interest from the spectrum data to obtain the error spectrum data. The spectrum data will show the frequency distribution of the error and help identify the characteristics of periodic errors.

[0087] Step S43: Perform an error peak and zero-crossing analysis on the error spectrum data to obtain error peak data; Specifically, by analyzing the peaks (local maximum points of the amplitude) and zero-crossing points (points where the amplitude is zero) in the spectrum data, important characteristics of the frequency components are identified. In the spectrum data, find each local maximum (peak). These peaks represent important periodic components in the frequency components. Identify the points where the amplitude is zero in the spectrum. These points correspond to changes or transitions in the frequency components. The zero-crossing points can be used to analyze the periodic characteristics of the error change. Through the peak and zero-crossing analysis, error peak data is obtained. These peaks and zero-crossing points will be used for further fuzzy clustering and error fusion feature extraction.

[0088] Step S44: Perform fuzzy clustering based on the error spectrum data and the error peak data to obtain error fusion feature data; Specifically, fuzzy clustering is a method of grouping data based on data similarity and is suitable for processing fuzzy, uncertain, or overlapping data. Combine the error spectrum data and the error peak data to form a multi-dimensional feature space, including information such as spectrum amplitude, frequency, and peak position. Calculate the similarity between each pair of data points. The similarity measure is based on information such as the difference in spectrum amplitude and the distance of peaks. Use algorithms such as fuzzy C-means to perform fuzzy clustering on the data. Each data point belongs to multiple categories with different membership degrees through calculation, thus obtaining a fuzzy classification result. Obtain the error fusion feature data. These data describe the error characteristics of different frequency components and can help understand the complex structure of the error.

[0089] Step S45: Perform a feature map conversion based on the error fusion feature data to obtain high-temperature eccentricity error feature data for assisting in the calibration operation of an eccentric turntable within a wide temperature range.

[0090] Specifically, the feature map conversion maps multi-dimensional feature data into a two-dimensional or three-dimensional space and presents it in a graphical way for easy analysis and use. The error fusion feature data is mapped into a two-dimensional coordinate system or a three-dimensional coordinate system, and visual elements such as color, size, and shape are used to represent the intensity, frequency, or other relevant information of the error. The error feature data is graphically represented to form a high-temperature eccentricity error feature map. These images can help analyze the spatial distribution, frequency characteristics, etc. of the error. The obtained high-temperature eccentricity error feature map can be used as a reference for subsequent eccentricity turntable calibration to assist in adjusting the eccentricity error of the turntable under different temperature conditions.

[0091] Preferably, the present application further provides a wide-temperature-range eccentricity turntable calibration system for performing the wide-temperature-range eccentricity turntable calibration method as described above. The wide-temperature-range eccentricity turntable calibration system includes: A high-temperature reference matrix data acquisition module for starting the eccentricity turntable in a high-temperature environment to obtain high-temperature reference matrix data; A high-temperature reference matrix region dissection module for performing region dissection according to the high-temperature reference matrix data to obtain high-temperature reference partition data; A deviation error change calculation module for calculating the eccentricity error according to the high-temperature reference partition data to obtain eccentricity error data, and performing error change processing according to the eccentricity error data to obtain deviation error change data; A high-temperature eccentricity error feature extraction module for constructing an error feature map according to the deviation error change data to obtain high-temperature eccentricity error feature data for assisting in the wide-temperature-range eccentricity turntable calibration operation.

[0092] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended application documents rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application documents are intended to be encompassed within the present invention.

[0093] The above description is only a specific implementation manner of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A calibration method for an eccentric turntable within a wide temperature range, characterized in that, It includes the following steps: Step S1: Start the eccentric turntable in a high-temperature environment to obtain high-temperature reference matrix data; Step S2: Perform regional dissection based on the high-temperature reference matrix data to obtain high-temperature reference partition data; Step S3: Calculate the eccentric error based on the high-temperature reference partition data to obtain eccentric error data, and perform error change processing based on the eccentric error data to obtain deviation error change data; Step S4: Construct an error characteristic map based on the deviation error change data to obtain high-temperature eccentric error characteristic data for assisting in the calibration of the eccentric turntable in a wide temperature range.

2. The method according to claim 1, wherein Specifically, Step S1 is as follows: Step S11: Start the eccentric turntable in a high-temperature environment, and obtain initial high-temperature environment perception data through sensors built in the eccentric turntable. The sensors built in the eccentric turntable include a high-temperature accelerometer, a fluxgate, and a gyroscope; Step S12: Perform temperature stratification modeling on the initial high-temperature environment perception data to obtain a temperature level model; Step S13: Perform multi-modal fusion based on the temperature level model to obtain high-temperature fusion data; Step S14: Use an inertial measurement unit to record the rotation trajectory, extract the rotation characteristics in a high-temperature environment, and obtain rotation trajectory data; Step S15: Construct a data trajectory based on the high-temperature fusion data and the rotation trajectory data to obtain high-temperature reference matrix data.

3. The method according to claim 1, characterized in that, Specifically, Step S2 is as follows: Step S21: Calculate the temperature gradient and vibration frequency based on the high-temperature reference matrix data to obtain temperature gradient data and vibration frequency data; Step S22: Perform temperature gradient region division on the high-temperature reference matrix data according to the temperature gradient data to obtain temperature gradient partition data; Step S23: Perform vibration frequency region division on the high-temperature reference matrix data according to the vibration frequency data to obtain vibration frequency partition data; Step S24: Perform change correlation analysis based on the temperature gradient data and the vibration frequency data to obtain temperature-vibration correlation data; Step S25: Map the temperature gradient partition data and the vibration frequency partition data according to the temperature-vibration correlation data to obtain temperature-vibration partition map data; Step S26: Perform error analysis and uncertainty analysis on the temperature-vibration partition map data to obtain partition map error data and partition map uncertainty data respectively; Step S27: Perform partition optimization on the temperature-vibration partition map data according to the partition map error data and the partition map uncertainty data to obtain the first high-temperature reference partition data; Step S28: Perform multi-scale regional dissection on the high-temperature reference matrix data to obtain the second high-temperature reference partition data.

4. The method according to claim 3, characterized in that Specifically, Step S21 is as follows: Perform temperature difference calculation based on the high-temperature reference matrix data to obtain temperature difference data; Perform temperature gradient calculation based on the temperature difference data to obtain temperature gradient data; Perform Fourier transform on the high-temperature reference matrix data to obtain frequency-domain vibration data; Extract the vibration frequency based on the frequency-domain vibration data to obtain vibration frequency data.

5. The method according to claim 3, characterized in that, Specifically, Step S22 is as follows: Construct a brick well thermodynamic model based on the temperature gradient data to obtain a brick well thermodynamic model; Extract the spatial variation characteristics of the temperature gradient according to the brick well thermodynamic model to obtain the temperature gradient spatial variation characteristic data; Set the temperature gradient threshold according to the temperature gradient spatial variation characteristic data to obtain the temperature gradient threshold data; Divide the high-temperature reference matrix data into temperature gradient regions through the temperature gradient threshold data to obtain the temperature gradient partition data.

6. The method according to claim 3, wherein Step S23 is specifically as follows: Perform wavelet transformation on the vibration frequency data to obtain the first vibration frequency characteristic data; Calculate the frequency change rate of the vibration frequency data according to the temperature gradient spatial variation characteristic data to obtain the second vibration frequency characteristic data; Divide the frequency change interval according to the first vibration frequency characteristic data and the second vibration frequency characteristic data to obtain the vibration frequency change interval data; Perform spatial region mapping according to the vibration frequency change interval data and the high-temperature reference matrix data to obtain the vibration frequency partition data.

7. The method according to claim 3, wherein Step S28 is specifically as follows: Perform non-linear image conversion on the high-temperature reference matrix data to obtain the high-temperature reference image data; Perform cross-scale spatial processing on the high-temperature reference image data to obtain the scale space data; Perform multi-scale blur processing on the scale space data through the preset Gaussian scale space parameter data to obtain the blurred scale image data; Calculate the scale gradient according to the blurred scale image data to obtain the scale gradient map data; Extract spatial features according to the scale gradient map data to obtain the scale space feature data; Perform clustering calculation on the scale space feature data to obtain the scale space region partition data; Perform data mapping on the high-temperature reference matrix data according to the scale space region partition data to obtain the second high-temperature reference partition data.

8. The method according to claim 1, wherein Step S3 is specifically as follows: Step S31: Analyze the local eccentricity effect according to the high-temperature reference partition data to obtain the local eccentricity effect data; Step S32: Construct a temperature vibration eccentricity error model according to the local eccentricity effect data to obtain the temperature vibration eccentricity error model; Step S33: Calculate the local error according to the high-temperature reference partition data and the temperature vibration eccentricity error model to obtain the eccentricity error data; Step S34: Analyze the time series error according to the eccentricity error data to obtain the deviation error change data.

9. The method according to claim 1, wherein Step S4 is specifically as follows: Step S41: Detrend the deviation error according to the deviation error change data to obtain the detrended error data; Step S42: Perform deviation error spectrum analysis on the detrended error data to obtain the error spectrum data; Step S43: Analyze the error peak and zero-crossing point of the error spectrum data to obtain the error peak data; Step S44: Perform fuzzy clustering according to the error spectrum data and the error peak data to obtain the error fusion characteristic data; Step S45: Perform feature map conversion according to the error fusion characteristic data to obtain the high-temperature eccentricity error characteristic data for assisting the calibration operation of the wide-temperature-range eccentric turntable.

10. An eccentric turntable calibration system with a wide temperature range, characterized in that For implementing the wide-temperature-range eccentric turntable calibration method as described in claim 1, the wide-temperature-range eccentric turntable calibration system includes: A high-temperature reference matrix data acquisition module for starting the eccentric turntable in a high-temperature environment to obtain high-temperature reference matrix data; A high-temperature reference matrix region partitioning module, which is used to perform region partitioning according to high-temperature reference matrix data to obtain high-temperature reference partition data; A bias error variation calculation module, which is used to calculate the eccentricity error according to the high-temperature reference partition data to obtain eccentricity error data, and perform error variation processing according to the eccentricity error data to obtain bias error variation data; A high-temperature eccentricity error feature extraction module, which is used to construct an error feature map according to the bias error variation data to obtain high-temperature eccentricity error feature data for assisting the calibration of an eccentric turntable in a wide temperature range.

Images