A mechanical plate surface roughness detection system

Abstract

The present invention relates to the field of surface roughness measurement technology, specifically a mechanical plate surface roughness detection system, which includes a stress distribution analysis module, a microelectronic force signal acquisition module, a data preprocessing module, a feature extraction and analysis module, a wear experiment control module, a spectral data processing module, and a comprehensive evaluation and output module. In the present invention, by applying the microelectronic force signal acquisition module, it is possible to accurately quantify the changes in tiny forces and surface layer stress, and to detect subtle changes in surface roughness at a more microscopic level. A machine learning algorithm is used to perform in-depth analysis of feature data, allowing the system to automatically identify key factors that affect surface roughness, thereby improving the intelligence level and adaptability of detection. The comprehensive evaluation and output module uses a multivariate regression analysis method, which not only quantifies the surface roughness of the mechanical plate, but also provides a new way to comprehensively evaluate the surface performance of the material, further improving the detection effect.

Description

Technical Field

[0001] The invention relates to the technical field of surface roughness measurement, and in particular to a mechanical plate surface roughness detection system. Background Art

[0002] Surface roughness measurement technology is a key area in precision engineering and manufacturing, involving the evaluation and quantification of the microstructure and texture of the material surface. Surface roughness directly affects the performance, life and reliability of mechanical parts, including reducing wear, improving fatigue strength, reducing friction and improving sealing effects. In modern manufacturing, with the increasing requirements for product quality and performance, the control of surface roughness has become particularly important, so the development of high-precision and high-efficiency surface roughness detection systems has become a research hotspot in this field.

[0003] Among them, the mechanical plate surface roughness detection system is a device specially designed to evaluate and measure the surface microstructure and roughness of mechanical plates. Its main purpose is to ensure that materials and products meet specific quality standards and performance requirements by accurately measuring surface roughness. By achieving accurate evaluation of surface roughness, it helps to improve product performance, extend service life, and ensure the stability and reliability of the manufacturing process. It can also be used for quality control, identification and correction of defects in the production process.

[0004] Traditional surface roughness measurement technology focuses on direct measurement or analysis based on a single data source, which limits the accuracy and comprehensiveness of detection. The lack of direct measurement and analysis of changes in the microscopic mechanical properties of materials makes it impossible to accurately identify the microscopic factors that affect surface roughness. A single data processing method cannot make full use of all available information, limiting the ability of the detection system to adapt to complex surface conditions. As a result, in application scenarios with high precision requirements, traditional technologies cannot provide sufficient detection accuracy and efficiency, affecting the control and optimization of product quality. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a mechanical plate surface roughness detection system.

[0006] In order to achieve the above-mentioned object, the present invention adopts the following technical scheme: a mechanical plate surface roughness detection system includes a stress distribution analysis module, a microelectronic force signal acquisition module, a data preprocessing module, a feature extraction and analysis module, a wear experiment control module, a spectral data processing module, and a comprehensive evaluation and output module;

[0007] The stress distribution analysis module simulates the shape of the mechanical plate when it is subjected to stress, evaluates the stress conditions of the difference areas, reveals the potential relationship between the stress concentration area and the surface roughness, and obtains the stress distribution map;

[0008] The microelectronic force signal acquisition module uses a piezoelectric sensor to measure surface stress, converts mechanical stress into electrical signals through the piezoelectric effect, accurately quantifies tiny forces and surface layer stress changes, and obtains stress electrical signal data;

[0009] The data preprocessing module performs noise filtering and signal normalization processing based on the stress distribution spectrum and the stress electrical signal data, optimizes the data quality, and obtains the preprocessed data;

[0010] The feature extraction and analysis module performs feature extraction and analysis based on the preprocessed data in combination with a machine learning algorithm, identifies surface roughness changes on a microscopic scale according to the analysis results, and obtains feature analysis results;

[0011] The wear test control module simulates a micro-wear process, measures the response of the material surface to the indentation load, evaluates the surface hardness and wear resistance, and obtains wear response data;

[0012] The spectral data processing module uses Raman spectroscopy technology to analyze changes in surface molecular vibration modes, reveal the chemical composition and structural information of the material surface, analyze the surface roughness, and generate chemical structure information;

[0013] The comprehensive evaluation and output module uses a multivariate regression analysis method based on feature analysis results, wear test data and chemical structure information to perform quantitative evaluation of the surface roughness of the mechanical plate and output a comprehensive evaluation record.

[0014] As a further solution of the present invention, the stress distribution map includes the direction of force, the range of influence and the maximum stress point, the stress electrical signal data includes the signal amplitude, frequency and change trend, the preprocessed data includes the signal after amplitude adjustment and the data set after removing outliers, the feature analysis results include the identified surface texture features, stress response characteristics and physical change indicators, the wear response data includes the indentation diameter, depth and wear rate measured in the experiment, the chemical structure information includes the surface element composition, chemical bond type and molecular vibration frequency, and the comprehensive evaluation record includes the comprehensive predicted roughness value, confidence interval and optimization suggestions.

[0015] As a further solution of the present invention, the stress distribution analysis module includes a morphology simulation submodule, a region assessment submodule, and a relationship disclosure submodule;

[0016] The morphology simulation submodule performs mechanical simulation based on the geometry and material properties of the mechanical plate, simulates the morphology change when subjected to force, and generates morphology change data;

[0017] The regional assessment submodule assesses the stress conditions of the different regions of the mechanical plate based on the morphological change data, identifies the regions with high stress, and obtains the stress region assessment results;

[0018] The relationship revealing submodule analyzes the potential relationship between the stress area and the surface roughness based on the stress area evaluation result, reveals the interaction between the two, and obtains the stress distribution map.

[0019] As a further solution of the present invention, the microelectronic force signal acquisition module includes a signal conversion submodule, a signal recording submodule, and a signal quantization submodule;

[0020] The signal conversion submodule measures the mechanical stress generated by the surface force using a piezoelectric sensor and converts it into an electrical signal through the piezoelectric effect to obtain a preliminary electrical signal;

[0021] The signal recording submodule records the change process of the electrical signal, including the waveform and frequency of the voltage and current, based on the preliminary electrical signal, and obtains the electrical signal record;

[0022] The signal quantification submodule analyzes and quantifies the electrical signal based on the electrical signal record, calculates the actual value of the tiny force and the surface layer stress change, and obtains the stress electrical signal data.

[0023] As a further solution of the present invention, the data preprocessing module includes a noise filtering submodule, a signal normalization submodule, and a data optimization submodule;

[0024] The noise filtering submodule uses digital filtering technology to remove background noise and interference based on the stress distribution map and stress electrical signal data to generate denoised data;

[0025] The signal normalization submodule performs maximum and minimum normalization processing based on the denoised data, unifies the data range into a specific interval, and obtains normalized data;

[0026] The data optimization submodule performs data quality assessment and optimization processing based on normalized data to ensure data consistency and reliability and obtain preprocessed data.

[0027] As a further solution of the present invention, the feature extraction and analysis module includes a data preparation submodule, a feature extraction submodule, and a roughness identification submodule;

[0028] The data preparation submodule formats the preprocessed data, including data type conversion and value range adjustment, to match the input requirements of the analysis tool to obtain analysis-adapted data;

[0029] The feature extraction submodule extracts key information from the analysis and adaptation data, including the peak value, mean value and variance of the data, integrates the parameters, and constructs a statistical feature set;

[0030] The roughness identification submodule uses the data in the statistical feature set and adopts a machine learning algorithm to perform in-depth analysis on the feature data, and judges and classifies the microscopic changes in surface roughness by setting thresholds and comparison standards to obtain feature analysis results;

[0031] The machine learning algorithm is based on the formula:

[0032]

[0033] Calculate the optimization problem and obtain the roughness classification results, where: is the normal vector of the hyperplane, is the bias term, is the square of the norm of the normal vector, is the regularization parameter, is the slack variable, n is the number of training samples, , , , is the weight coefficient, is the surface wavelength ratio, is the peak density, is the surface inclination, is the morphological complexity.

[0034] As a further solution of the present invention, the wear experiment control module includes a wear simulation submodule, an indentation response measurement submodule, and a hardness and wear resistance evaluation submodule;

[0035] The wear simulation submodule simulates the micro-wear process, uses a controlled force to perform an indentation experiment on the material surface, simulates the wear conditions in actual use, and obtains wear simulation data;

[0036] The indentation response measurement submodule measures the response of the material surface to the indentation load based on the wear simulation data, records the indentation depth and width data, and obtains the indentation response data;

[0037] The hardness and wear resistance evaluation submodule evaluates the surface hardness and wear resistance of the material based on the indentation response data, analyzes the wear resistance of the material, and obtains the wear response data.

[0038] As a further solution of the present invention, the spectral data processing module includes a Raman spectrum acquisition submodule, a molecular vibration analysis submodule, and a chemical structure analysis submodule;

[0039] The Raman spectrum acquisition submodule performs Raman spectrum scanning on the surface of the material, collects scattering spectrum data caused by surface molecular vibration, and obtains a scattering spectrum data set;

[0040] The molecular vibration analysis submodule analyzes the Raman scattering peaks corresponding to the differential molecular vibration modes based on the scattering spectrum data set, identifies the key chemical components on the material surface, and generates vibration mode analysis results;

[0041] The chemical structure analysis submodule utilizes the vibration mode analysis results to integrate the types of chemical bonds and molecular structures, reveal the chemical composition and structural information of the material surface, analyze the surface roughness, and obtain chemical structure information.

[0042] As a further solution of the present invention, the comprehensive evaluation and output module includes a data integration submodule, a regression analysis submodule, and an evaluation result output submodule;

[0043] The data integration submodule integrates the characteristic analysis results, wear test data and chemical structure information, unifies the data format and scale, including the test results and all parameters required for analysis, and constructs a comprehensive analysis data set;

[0044] The regression analysis submodule performs multivariate regression analysis on the data in the comprehensive analysis data set, assigns weights according to the contribution of factors to surface roughness, sets parameters, executes the regression analysis process, reveals the key factors affecting roughness, and generates roughness influencing factor analysis results;

[0045] The multivariate regression analysis was performed according to the formula:

[0046]

[0047] Perform weight assignment, set parameters, perform regression analysis, and generate roughness influencing factor analysis results, among which: is the surface roughness, is the intercept term, , ,......, is the coefficient of the traditional independent variable, , ,......, is the traditional predictor variable of influence, is the weight of the material hardness coefficient, is the material hardness coefficient, is the weight of the cutting speed influence coefficient, is the cutting speed influence coefficient, is the weight of coolant usage, For coolant usage, is the weight of tool wear degree, is the degree of tool wear, is the error term;

[0048] The evaluation result output submodule utilizes the roughness influencing factor analysis results to quantify the influence of each factor on the surface roughness of the mechanical plate, integrates the quantified results, and obtains a comprehensive evaluation record.

[0049] Compared with the prior art, the advantages and positive effects of the present invention are:

[0050] In the present invention, by applying the microelectronic force signal acquisition module, the tiny forces and surface layer stress changes can be accurately quantified, and the subtle changes in surface roughness can be detected at a more microscopic level. The machine learning algorithm is used to perform in-depth analysis of the characteristic data, allowing the system to automatically identify the key factors affecting the surface roughness, thereby improving the intelligence level and adaptability of the detection. The comprehensive evaluation and output module uses a multivariate regression analysis method, which not only quantifies the surface roughness of the mechanical plate, but also provides a new way to comprehensively evaluate the surface performance of the material, further improving the detection effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a system flow chart of the present invention;

[0052] Figure 2 It is a schematic diagram of the system framework of the present invention;

[0053] Figure 3 It is a flow chart of the stress distribution analysis module of the present invention;

[0054] Figure 4 This is a flow chart of the microelectronic force signal acquisition module of the present invention;

[0055] Figure 5 It is a flow chart of the data preprocessing module of the present invention;

[0056] Figure 6 It is a flow chart of the feature extraction and analysis module of the present invention;

[0057] Figure 7 This is a flow chart of the wear test control module of the present invention;

[0058] Figure 8 is a flow chart of the spectrum data processing module of the present invention;

[0059] Fig. 9 It is a flow chart of the comprehensive evaluation and output module of the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0061] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0062] Embodiment 1

[0063] See also Figure 1 and Figure 2 ,A mechanical plate surface roughness detection system includes a stress distribution analysis module, a microelectronic force signal acquisition module, a data preprocessing module, a feature extraction and analysis module, a wear experiment control module, a spectrum data processing module, and a comprehensive evaluation and output module;

[0064] The stress distribution analysis module simulates the shape of the mechanical plate when it is subjected to stress, evaluates the stress conditions in different areas, reveals the potential relationship between stress concentration areas and surface roughness, and obtains stress distribution maps;

[0065] The microelectronic force signal acquisition module uses piezoelectric sensors to measure surface stress, converts mechanical stress into electrical signals through the piezoelectric effect, accurately quantifies tiny forces and surface stress changes, and obtains stress electrical signal data;

[0066] The data preprocessing module performs noise filtering and signal normalization based on the stress distribution spectrum and stress electrical signal data to optimize data quality and obtain preprocessed data;

[0067] The feature extraction and analysis module extracts and analyzes features based on the preprocessed data in combination with machine learning algorithms, identifies surface roughness changes at the microscopic scale based on the analysis results, and obtains feature analysis results;

[0068] The wear test control module simulates the micro-wear process, measures the response of the material surface to the indentation load, evaluates the surface hardness and wear resistance, and obtains the wear response data;

[0069] The spectral data processing module uses Raman spectroscopy technology to analyze the changes in the vibration modes of surface molecules, reveal the chemical composition and structural information of the material surface, analyze the surface roughness, and generate chemical structure information;

[0070] The comprehensive evaluation and output module uses multivariate regression analysis methods based on feature analysis results, wear test data and chemical structure information to conduct quantitative evaluation of the surface roughness of mechanical plates and output comprehensive evaluation records.

[0071] The stress distribution map includes the direction of force, the range of influence and the maximum stress point. The stress electrical signal data includes the signal amplitude, frequency and change trend. The preprocessed data includes the signal after amplitude adjustment and the data set after removing outliers. The feature analysis results include the identified surface texture features, stress response characteristics and physical change indicators. The wear response data includes the indentation diameter, depth and wear rate measured in the experiment. The chemical structure information includes the surface element composition, chemical bond type and molecular vibration frequency. The comprehensive evaluation record includes the comprehensive predicted roughness value, confidence interval and optimization suggestions.

[0072] See also Figure 2 and Figure 3 ,The stress distribution analysis module includes the morphological simulation submodule, the regional assessment submodule, and the relationship revealing submodule;

[0073] The morphology simulation submodule performs mechanical simulation based on the geometry and material properties of the mechanical plate, simulates the morphology changes when subjected to force, and generates morphology change data;

[0074] In the morphological simulation submodule, based on the geometric properties and material characteristics of the mechanical plate, the finite element analysis software ANSYS is used, and its built-in static analysis function is used to set boundary conditions to simulate the stress conditions in the actual working environment. The mechanical plate is meshed and refined to the microscopic scale. Parameters such as the elastic modulus and Poisson's ratio of the material are specified, and the mechanical simulation process is executed. The morphological change data of the mechanical plate when subjected to stress is generated, and the displacement of each mesh node during the stress process is recorded in detail.

[0075] The regional assessment submodule evaluates the stress conditions of the mechanical plate's different regions based on the morphological change data, identifies the regions with high stress, and obtains the stress region assessment results;

[0076] In the regional assessment submodule, based on the morphological change data, Python programming is used and the NumPy library is used for data processing. By comparing the displacement of each node, the size and distribution area of ​​the deformation after force are calculated, and a threshold is set to identify the areas with greater force. These areas are marked and classified, and the position and deformation of the areas with greater force are recorded to obtain the assessment results of the force area, which includes the specific position and force value of the identified high-force area.

[0077] The relationship revealing submodule analyzes the potential relationship between the stress area and the surface roughness based on the stress area assessment results, reveals the interaction between the two, and obtains the stress distribution map;

[0078] In the relationship revealing submodule, based on the evaluation results of the stress area, the Matplotlib library is used for data visualization, and a force distribution diagram is drawn. The force magnitude is represented by the depth of color. The linear regression model in the Scikit-learn library is used to analyze the relationship between the stress area and the surface roughness. The independent variable of the linear regression model is set as the force magnitude, and the dependent variable is set as the roughness measurement value of the corresponding area. Linear regression analysis is performed to reveal the linear relationship between the force magnitude and the surface roughness, and a stress distribution map is generated to show the interaction relationship between the stress area and the surface roughness.

[0079] See also Figure 2 and Figure 4 ,The microelectronic force signal acquisition module includes a signal conversion submodule, a signal recording submodule, and a signal quantization submodule;

[0080] The signal conversion submodule uses a piezoelectric sensor to measure the mechanical stress generated by the surface force, and converts it into an electrical signal through the piezoelectric effect to obtain a preliminary electrical signal;

[0081] In the signal conversion submodule, a piezoelectric sensor is used to measure the mechanical stress generated by the surface force, a charge amplifier is used to amplify the tiny charge signal generated by the piezoelectric effect, and an analog-to-digital converter (ADC) is used to convert the amplified analog signal into a digital signal. The accuracy of the signal conversion is adjusted by setting the amplification factor of the charge amplifier and the sampling frequency of the ADC to ensure that tiny mechanical changes are captured, thereby generating a preliminary electrical signal.

[0082] The signal recording submodule records the change process of the electrical signal based on the preliminary electrical signal, including the waveform and frequency of the voltage and current, and obtains the electrical signal record;

[0083] In the signal recording submodule, based on the preliminary electrical signal, the data acquisition system (DAQ) is used to record the change process of the electrical signal. The sampling rate of the DAQ is set to ensure sufficient data resolution. At the same time, a low-pass filter is used to filter out high-frequency noise to ensure that the recorded electrical signal clearly reflects the changes in the electrical signal caused by the force, including the waveforms of voltage and current and their frequency changes, to obtain electrical signal records.

[0084] The signal quantification submodule analyzes and quantifies the electrical signals based on the electrical signal records, calculates the actual values ​​of the tiny forces and surface stress changes, and obtains stress electrical signal data;

[0085] In the signal quantification submodule, based on the electrical signal recording, the signal processing software is used to perform fast Fourier transform (FFT) on the electrical signal, analyze the spectral characteristics of the signal, and convert the signal intensity into force value according to the preset force and signal intensity relationship model. By setting the parameters in the conversion model, such as the proportional coefficient of signal intensity and force, the actual values ​​of the tiny force caused by the surface force and the stress change in the surface layer are accurately calculated, thereby obtaining the stress electrical signal data.

[0086] See also Figure 2 and Figure 5 ,The data preprocessing module includes a noise filtering submodule, a signal normalization submodule, and a data optimization submodule;

[0087] The noise filtering submodule uses digital filtering technology to remove background noise and interference based on stress distribution maps and stress electrical signal data to generate denoised data;

[0088] In the noise filtering submodule, based on the stress distribution map and stress electrical signal data, the signal module in the SciPy library in the Python environment is used, and the Butterworth low-pass filter is used. The filter order is set to 4 and the cutoff frequency is set to 150 Hz. The purpose is to remove background noise and high-frequency interference, perform the filtering process, ensure the smoothness and accuracy of the signal, and thus generate denoised data.

[0089] The signal normalization submodule performs maximum and minimum normalization processing based on the denoised data, unifies the data range into a specific interval, and obtains normalized data;

[0090] In the signal normalization submodule, based on the denoised data, the NumPy library in the Python environment is used to perform maximum and minimum normalization processing. By calculating the maximum and minimum values ​​in the data set, and subtracting the minimum value from each value and dividing it by the difference between the maximum and minimum values, all data are adjusted to the [0,1] interval. This process aims to unify the data ranges of different magnitudes, ensure the accuracy of subsequent analysis, and obtain normalized data.

[0091] The data optimization submodule performs data quality assessment and optimization based on normalized data to ensure data consistency and reliability and obtain preprocessed data;

[0092] In the data optimization submodule, a data consistency check method is adopted based on normalized data. Here, the Pandas library of the Python environment is used to traverse the data set to check whether each data point meets the pre-defined consistency standards, such as numerical type and range, and exclude any outliers or data points with large deviations. This step aims to further improve data quality and ensure the reliability of analysis results, thereby obtaining preprocessed data.

[0093] See also Figure 2 and Figure 6 ,The feature extraction and analysis module includes a data preparation submodule, a feature extraction submodule, and a roughness recognition submodule;

[0094] The data preparation submodule formats the preprocessed data, including data type conversion and value range adjustment, to match the input requirements of the analysis tool and obtain analysis-adapted data;

[0095] In the data preparation submodule, based on the preprocessed data, Python scripts are used to format the data, and the Pandas library is used to convert the data type to floating point numbers to ensure the compatibility of subsequent calculations. The data range is adjusted, and the MinMaxScaler function of the Sklearn library is used for maximum and minimum normalization. This function adjusts the data range to [0,1] by calculating the maximum and minimum values ​​of the original data, ensuring that data from different sources are in the same numerical range, thereby obtaining analysis-adapted data.

[0096] The feature extraction submodule extracts key information from the analyzed adaptation data, including the peak value, mean value, and variance of the data, integrates the parameters, and constructs a statistical feature set;

[0097] In the feature extraction submodule, based on the analysis of the adaptation data, Python scripts are used to extract key information using the NumPy library, and the statistical parameters of the data set are calculated, including the peak value, mean, standard deviation, and variance. The statistical parameters provide quantitative characteristic indicators for subsequent roughness analysis by accurately calculating the relationship between each data point and the overall data set. The statistical parameters are integrated into a feature vector to construct a statistical feature set.

[0098] The roughness identification submodule uses the data in the statistical feature set and a machine learning algorithm to conduct in-depth analysis of the feature data. It determines and classifies the microscopic changes in surface roughness by setting thresholds and comparison standards to obtain feature analysis results.

[0099] Machine learning algorithm, according to the formula:

[0100]

[0101] Calculate the optimization problem and obtain the roughness classification results, where: is the normal vector of the hyperplane, which is used to determine the direction of the optimal segmentation surface. is the bias term, adjusting the position of the hyperplane, is the square of the norm of the normal vector. Minimizing this value can maximize the classification interval. is a regularization parameter used to balance the complexity of the model and the fit of the training data. is a slack variable that allows data points to violate the interval boundary to a certain extent, n is the number of training samples, , , , is the weight coefficient, which is used to adjust the intensity of the impact of the new parameters on the model. is the surface wavelength ratio, reflecting the coarseness of the surface texture. is the peak density, the ratio of the number of peaks on the surface to the total measured area, is the surface inclination, the average absolute value of the first derivative of the surface height function, For morphological complexity, it provides information on the diversity of surface structures.

[0102] The execution process is as follows:

[0103] Feature extraction and parameter calculation: collect data on the surface of the object, extract basic statistical features from it, and calculate new surface characteristic parameters ( , , , );

[0104] Model training and optimization, taking the extracted features and calculated parameters as input, training the SVM model, adjusting , and determine , and , , , The value of , and then use cross validation and grid search techniques to determine the optimal value and weight coefficient ( , , , ), ensuring that the model has both good generalization ability and accurate classification;

[0105] Classification and evaluation: Use the trained model to classify the newly collected data in roughness, use evaluation indicators (such as accuracy and recall) to evaluate the performance of the model, and adjust the model parameters as needed.

[0106] Through this method, the improved machine learning algorithm can comprehensively utilize traditional statistical features and newly added surface characteristic parameters, and achieve more accurate identification and classification of the surface roughness of objects by precisely adjusting the model parameters, thereby improving the accuracy and efficiency of roughness identification and providing strong technical support for precision surface engineering.

[0107] See also Figure 2 and Figure 7,The wear experiment control module includes the wear simulation submodule, the indentation response measurement submodule, and the hardness and wear resistance evaluation submodule;

[0108] The wear simulation submodule simulates the micro-wear process, uses controlled force to perform indentation experiments on the material surface, simulates the wear conditions in actual use, and obtains wear simulation data;

[0109] In the wear simulation submodule, an indentation experiment is performed on the material surface based on the controlled force. A dedicated material testing machine, such as a universal material testing machine, is used. The loading rate of the testing machine is set to 1N / s and the loading depth is controlled within 0.1mm to simulate the slight wear that may be encountered in actual use. The loading force is recorded by the built-in force sensor of the testing machine, and the indentation depth is recorded by a displacement sensor to obtain wear simulation data.

[0110] The indentation response measurement submodule measures the response of the material surface to the indentation load based on the wear simulation data, records the indentation depth and width data, and obtains the indentation response data;

[0111] In the indentation response measurement submodule, based on the wear simulation data, a microscope is used in conjunction with image analysis software, such as ImageJ, to observe and measure the indentation. The scale calibration parameters in the image analysis software are set to ensure measurement accuracy. The software automatically identifies the indentation edge, calculates the maximum depth and width of the indentation, and records these data as the response of the material surface to the indentation load to obtain the indentation response data.

[0112] The hardness and wear resistance evaluation submodule evaluates the surface hardness and wear resistance of the material based on the indentation response data, analyzes the wear resistance of the material, and obtains the wear response data;

[0113] In the hardness and wear resistance evaluation submodule, the hardness of the material is evaluated based on the indentation response data using a hardness calculation formula, such as the Vickers hardness calculation formula. The hardness calculation formula takes into account the relationship between the indentation load and the indentation area. By setting the parameters in the formula, such as the loading force and the indentation area, the hardness value of the material is calculated. At the same time, the wear resistance of the material is evaluated based on the indentation size, and the wear resistance of the material is analyzed to obtain the wear response data.

[0114] See also Figure 2 and Figure 8 ,The spectral data processing module includes a Raman spectrum acquisition submodule, a molecular vibration analysis submodule, and a chemical structure analysis submodule;

[0115] The Raman spectrum acquisition submodule performs Raman spectrum scanning on the material surface, collects the scattering spectrum data caused by the surface molecular vibration, and obtains the scattering spectrum data set;

[0116] In the Raman spectrum acquisition submodule, Raman spectrum scanning is performed based on the material surface. A Raman spectrum instrument is used, the laser wavelength is set to 532nm, the sampling interval is 1cm^-1, and the scanning range is set to 400-4000cm^-1 to cover most typical Raman scattering peaks of organic and inorganic substances. The material surface is excited by laser, and the scattering spectrum signal caused by the surface molecular vibration is collected. The signal is preliminarily processed using the software provided by the instrument, including background noise removal and signal baseline correction, to obtain a scattering spectrum data set.

[0117] The molecular vibration analysis submodule analyzes the Raman scattering peaks corresponding to the differential molecular vibration modes based on the scattering spectrum data set, identifies the key chemical components on the material surface, and generates vibration mode analysis results;

[0118] In the molecular vibration analysis submodule, based on the scattering spectrum data set, spectral analysis software such as Origin or SpectraGryph is used to import the data set for further analysis. By setting the search peak algorithm parameters, such as the minimum peak width and signal-to-noise ratio threshold, Raman scattering peaks are automatically identified and marked. The scattering peaks correspond to different molecular vibration modes. By comparing with known Raman scattering databases, such as the RRUFF database, the key chemical components on the material surface are identified and vibration mode analysis results are generated.

[0119] The chemical structure analysis submodule uses the vibration mode analysis results to integrate the type of chemical bonds and molecular structure, reveal the chemical composition and structural information of the material surface, analyze the surface roughness, and obtain chemical structure information;

[0120] In the chemical structure analysis submodule, the vibration mode analysis results are used, chemical analysis software is used, and theoretical knowledge of chemical bond types and molecular structures are combined to further analyze the identified chemical components and analyze their distribution and interactions on the material surface. The structural simulation function in the software is used to infer the chemical composition and structural information of the material surface based on the intensity and distribution of the Raman scattering peak, and the influence of surface roughness is qualitatively analyzed to obtain chemical structure information.

[0121] See also Figure 2 and Fig. 9 ,The comprehensive evaluation and output module includes a data integration submodule, a regression analysis submodule, and an evaluation result output submodule;

[0122] The data integration submodule integrates the feature analysis results, wear test data and chemical structure information, unifies the data format and scale, including the test results and all parameters required for analysis, and constructs a comprehensive analysis data set;

[0123] In the data integration submodule, Python scripts and Pandas library are used to merge data and unify formats based on feature analysis results, wear test data and chemical structure information. First, the data set is cleaned to remove missing values ​​and outliers. Then, the merge function of Pandas is used to merge data sets from different sources based on a common identifier. The MinMaxScaler function is then used to adjust the numerical range of the merged data set to ensure that all data are on the same scale, thereby constructing a comprehensive analysis data set.

[0124] The regression analysis submodule performs multivariate regression analysis on the data in the comprehensive analysis data set, assigns weights based on the contribution of factors to surface roughness, sets parameters, executes the regression analysis process, reveals the key factors affecting roughness, and generates analysis results of factors affecting roughness;

[0125] Multivariate regression analysis was performed according to the formula:

[0126]

[0127] Perform weight assignment, set parameters, perform regression analysis, and generate roughness influencing factor analysis results, among which: is the response variable, specifically surface roughness, is the intercept term, which means when all independent variables are 0 The expected value of , ,......, is the coefficient of the traditional independent variable, indicating the effect of each factor on The intensity of the impact, , ,......, is the traditional predictor variable of influence, is the weight of the material hardness coefficient, is the material hardness coefficient, is the weight of the cutting speed influence coefficient, is the cutting speed influence coefficient, is the weight of coolant usage, For coolant usage, is the weight of tool wear degree, is the degree of tool wear, is the error term, which represents the random variation that cannot be explained by the model.

[0128] The execution process is as follows:

[0129] Data preparation, collection including surface roughness , traditional influencing factors ( , , ...) and the creative parameters ( , , , ) data;

[0130] Model construction, based on multivariate regression analysis, adding material hardness coefficient , cutting speed influence coefficient , Coolant usage and tool wear as new predictor variables;

[0131] Parameter estimation, using extended least squares or other appropriate statistical methods to estimate the intercept term , the coefficients of the traditional independent variables ( , , ...) and the coefficients of the innovation parameters ( , , , );

[0132] Model evaluation and validation, using statistical tests to assess the overall fit of the model and the significance of each predictor variable;

[0133] Coefficient interpretation: Based on the estimated coefficient values, explain the degree and direction of influence of various factors (including the newly introduced creative factors) on surface roughness;

[0134] Application of results: The model is used to predict and control surface roughness, providing guidance and optimization suggestions for the manufacturing process.

[0135] By changing the material hardness coefficient , cutting speed influence coefficient , Coolant usage , Tool wear degree The introduction of four creative parameters into the multivariate regression model not only increases the complexity of the model, but also improves its adaptability to real-world conditions and prediction accuracy. It can more accurately generate analysis results of factors affecting surface roughness and provide a more precise basis for engineering design and optimization of production processes.

[0136] The evaluation result output submodule uses the roughness influencing factor analysis results to quantify the impact of each factor on the surface roughness of the mechanical board, integrates the quantified results, and obtains a comprehensive evaluation record;

[0137] In the evaluation result output submodule, the roughness influencing factor analysis results are used to quantify and integrate the results using Python scripts. First, the coefficients of the regression analysis model are extracted as the weights of the influence of each factor on the surface roughness. Then, the specific contribution of each factor to the surface roughness is calculated based on the weights. The Matplotlib library is used to visualize the results and a bar chart of the influence of each factor on the surface roughness is drawn. Finally, all the analysis results are summarized into a document and output in PDF format to ensure the readability and dissemination of the results, thereby obtaining a comprehensive evaluation record.

[0138] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A mechanical board surface roughness detection system, characterized in that: The mechanical plate surface roughness detection system includes a stress distribution analysis module, a microelectronic force signal acquisition module, a data preprocessing module, a feature extraction and analysis module, a wear experiment control module, a spectrum data processing module, and a comprehensive evaluation and output module; The stress distribution analysis module simulates the shape of the mechanical plate when it is subjected to stress, evaluates the stress conditions of the difference areas, reveals the potential relationship between the stress concentration area and the surface roughness, and obtains the stress distribution map; The microelectronic force signal acquisition module uses a piezoelectric sensor to measure surface stress, converts mechanical stress into electrical signals through the piezoelectric effect, accurately quantifies tiny forces and surface layer stress changes, and obtains stress electrical signal data; The data preprocessing module performs noise filtering and signal normalization processing based on the stress distribution spectrum and the stress electrical signal data, optimizes the data quality, and obtains the preprocessed data; The feature extraction and analysis module performs feature extraction and analysis based on the preprocessed data in combination with a machine learning algorithm, identifies surface roughness changes on a microscopic scale according to the analysis results, and obtains feature analysis results; The wear test control module simulates a micro-wear process, measures the response of the material surface to the indentation load, evaluates the surface hardness and wear resistance, and obtains wear response data; The spectral data processing module uses Raman spectroscopy technology to analyze changes in surface molecular vibration modes, reveal the chemical composition and structural information of the material surface, analyze the surface roughness, and generate chemical structure information; The comprehensive evaluation and output module uses a multivariate regression analysis method based on feature analysis results, wear test data and chemical structure information to perform quantitative evaluation of the surface roughness of the mechanical plate and output a comprehensive evaluation record.

2. The mechanical board surface roughness detection system according to claim 1, characterized in that: The stress distribution map includes the direction of force, the range of influence and the maximum stress point, the stress electrical signal data includes the signal amplitude, frequency and change trend, the preprocessed data includes the signal after amplitude adjustment and the data set after removing outliers, the feature analysis results include the identified surface texture features, stress response characteristics and physical change indicators, the wear response data includes the indentation diameter, depth and wear rate measured in the experiment, the chemical structure information includes the surface element composition, chemical bond type and molecular vibration frequency, and the comprehensive evaluation record includes the comprehensive predicted roughness value, confidence interval and optimization suggestions.

3. The mechanical board surface roughness detection system according to claim 1, characterized in that: The stress distribution analysis module includes a morphology simulation submodule, a region assessment submodule, and a relationship disclosure submodule; The morphology simulation submodule performs mechanical simulation based on the geometry and material properties of the mechanical plate, simulates the morphology change when subjected to force, and generates morphology change data; The regional assessment submodule assesses the stress conditions of the different regions of the mechanical plate based on the morphological change data, identifies the regions with high stress, and obtains the stress region assessment results; The relationship revealing submodule analyzes the potential relationship between the stress area and the surface roughness based on the stress area evaluation result, reveals the interaction between the two, and obtains the stress distribution map.

4. The mechanical board surface roughness detection system according to claim 1, characterized in that: The microelectronic force signal acquisition module includes a signal conversion submodule, a signal recording submodule, and a signal quantization submodule; The signal conversion submodule measures the mechanical stress generated by the surface force using a piezoelectric sensor and converts it into an electrical signal through the piezoelectric effect to obtain a preliminary electrical signal; The signal recording submodule records the change process of the electrical signal, including the waveform and frequency of the voltage and current, based on the preliminary electrical signal, and obtains the electrical signal record; The signal quantification submodule analyzes and quantifies the electrical signal based on the electrical signal record, calculates the actual value of the tiny force and the surface layer stress change, and obtains the stress electrical signal data.

5. The mechanical board surface roughness detection system according to claim 1, characterized in that: The data preprocessing module includes a noise filtering submodule, a signal normalization submodule, and a data optimization submodule; The noise filtering submodule uses digital filtering technology to remove background noise and interference based on the stress distribution map and stress electrical signal data to generate denoised data; The signal normalization submodule performs maximum and minimum normalization processing based on the denoised data, unifies the data range into a specific interval, and obtains normalized data; The data optimization submodule performs data quality assessment and optimization processing based on normalized data to ensure data consistency and reliability and obtain preprocessed data.

6. The mechanical board surface roughness detection system according to claim 1, characterized in that: The feature extraction and analysis module includes a data preparation submodule, a feature extraction submodule, and a roughness identification submodule; The data preparation submodule formats the preprocessed data, including data type conversion and value range adjustment, to match the input requirements of the analysis tool to obtain analysis-adapted data; The feature extraction submodule extracts key information from the analysis and adaptation data, including the peak value, mean value and variance of the data, integrates the parameters, and constructs a statistical feature set; The roughness identification submodule uses the data in the statistical feature set and adopts a machine learning algorithm to perform in-depth analysis on the feature data, and judges and classifies the microscopic changes in surface roughness by setting thresholds and comparison standards to obtain feature analysis results; The machine learning algorithm is based on the formula: ； Calculate the optimization problem and obtain the roughness classification results, where: is the normal vector of the hyperplane, is the bias term, is the square of the norm of the normal vector, is the regularization parameter, is the slack variable, n is the number of training samples, , , , is the weight coefficient, is the surface wavelength ratio, is the peak density, is the surface inclination, is the morphological complexity.

7. The mechanical board surface roughness detection system according to claim 1, characterized in that: The wear experiment control module includes a wear simulation submodule, an indentation response measurement submodule, and a hardness and wear resistance evaluation submodule; The wear simulation submodule simulates the micro-wear process, uses a controlled force to perform an indentation experiment on the material surface, simulates the wear conditions in actual use, and obtains wear simulation data; The indentation response measurement submodule measures the response of the material surface to the indentation load based on the wear simulation data, records the indentation depth and width data, and obtains the indentation response data; The hardness and wear resistance evaluation submodule evaluates the surface hardness and wear resistance of the material based on the indentation response data, analyzes the wear resistance of the material, and obtains the wear response data.

8. The mechanical board surface roughness detection system according to claim 1, characterized in that: The spectral data processing module includes a Raman spectrum acquisition submodule, a molecular vibration analysis submodule, and a chemical structure analysis submodule; The Raman spectrum acquisition submodule performs Raman spectrum scanning on the surface of the material, collects scattering spectrum data caused by surface molecular vibration, and obtains a scattering spectrum data set; The molecular vibration analysis submodule analyzes the Raman scattering peaks corresponding to the differential molecular vibration modes based on the scattering spectrum data set, identifies the key chemical components on the material surface, and generates vibration mode analysis results; The chemical structure analysis submodule utilizes the vibration mode analysis results to integrate the types of chemical bonds and molecular structures, reveal the chemical composition and structural information of the material surface, analyze the surface roughness, and obtain chemical structure information.

9. The mechanical board surface roughness detection system according to claim 1, characterized in that: The comprehensive evaluation and output module includes a data integration submodule, a regression analysis submodule, and an evaluation result output submodule; The data integration submodule integrates the characteristic analysis results, wear test data and chemical structure information, unifies the data format and scale, including the test results and all parameters required for analysis, and constructs a comprehensive analysis data set; The regression analysis submodule performs multivariate regression analysis on the data in the comprehensive analysis data set, assigns weights according to the contribution of factors to surface roughness, sets parameters, executes the regression analysis process, reveals the key factors affecting roughness, and generates roughness influencing factor analysis results; The multivariate regression analysis was performed according to the formula: ； Perform weight assignment, set parameters, perform regression analysis, and generate roughness influencing factor analysis results, among which: is the surface roughness, is the intercept term, , ,......, is the coefficient of the traditional independent variable, , ,......, is the traditional predictor variable of influence, is the weight of the material hardness coefficient, is the material hardness coefficient, is the weight of the cutting speed influence coefficient, is the cutting speed influence coefficient, is the weight of coolant usage, For coolant usage, is the weight of tool wear degree, is the degree of tool wear, is the error term; The evaluation result output submodule utilizes the roughness influencing factor analysis results to quantify the influence of each factor on the surface roughness of the mechanical plate, integrates the quantified results, and obtains a comprehensive evaluation record.

Images