Method for actively controlling slag deposition state of cutter cylinder and inhibiting abrasion of cutter

Abstract

The invention relates to the technical field of machining, and discloses a cutter cylinder slag deposition state active control and cutter wear inhibition method, which is applied to a machining control system, and the system comprises a plurality of slag deposition monitoring sensors arranged on the inner wall of a cutter cylinder and an active control module. The method comprises the following steps that vibration spectrum data and temperature gradient data are collected through a slag deposition monitoring sensor, and a slag deposition state data set is generated; the active control module receives real-time data to construct a slag deposition state matrix; determining spatio-temporal evolution characteristics of slag deposition: processing a slag deposition state matrix, extracting characteristics, modeling, predicting evolution, constructing an active control model, outputting spatio-temporal evolution characteristics, and updating a data set; and dynamically regulating and controlling the cutter cylinder cooling medium according to the evolution characteristics, wherein the dynamic regulation and control comprise mapping section identification, controlling execution action and dynamically distributing the cooling medium. According to the method, precise monitoring and active control over the slag deposition state of the cutter cylinder are achieved, cutter abrasion is effectively restrained, the machining quality and efficiency are improved, and the service life of a cutter and equipment is prolonged.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mechanical processing, specifically a tool barrel accumulation state active control and tool wear inhibition method. BACKGROUND

[0002] In the field of mechanical processing, the tool barrel as a key component, its accumulation state and tool wear have a significant impact on processing quality, efficiency and equipment life. The traditional tool barrel accumulation control and tool wear inhibition method has many limitations.

[0003] From the perspective of accumulation control, in the early mechanical processing, the monitoring and handling of tool barrel accumulation were relatively passive. The operator mainly relied on experience to judge the accumulation situation, such as observing abnormal vibration, sound during processing or regularly stopping to check the accumulation state of the inner wall of the tool barrel. This method is not only inefficient, but also cannot accurately grasp the real-time state of the accumulation, which easily leads to excessive deposition of accumulation and affects the normal operation of the tool barrel. With the development of technology, although some monitoring methods based on simple sensors have appeared, such as using pressure sensors or temperature sensors to monitor the physical parameters inside the tool barrel, these single type sensors can only obtain limited information and cannot fully reflect the complex state of the accumulation. For example, the pressure sensor can only monitor the pressure change, but cannot directly reflect the deposition position and growth trend of the accumulation; the temperature sensor can sense the temperature change, but has no effect on the vibration characteristics of the accumulation. In addition, the traditional accumulation control method often uses a fixed cooling medium regulation strategy, which cannot be dynamically adjusted according to the real-time evolution characteristics of the accumulation, resulting in low efficiency of the use of cooling medium and inability to effectively inhibit the formation and development of the accumulation.

[0004] In terms of tool wear inhibition, tool wear is an inevitable problem in mechanical processing, and unreasonable control of tool barrel accumulation will further exacerbate tool wear. The traditional tool wear inhibition method mainly focuses on the improvement of tool material and the optimization of tool structure, such as using tool materials with better wear resistance or designing more reasonable tool geometry. However, these methods, although to some extent, can prolong the service life of the tool, ignore the internal relationship between tool barrel accumulation and tool wear. The uneven deposition of accumulation in the tool barrel will cause the tool to bear uneven load and friction during processing, thereby accelerating the wear of the tool. At the same time, the traditional method lacks in-depth study of the correlation between the accumulation state and tool wear, and cannot effectively inhibit the tool wear by actively controlling the accumulation state.

[0005] With the development of the mechanical processing industry towards high precision, high efficiency and automation, higher requirements are put forward for the control of the accumulation state of the tool barrel and the inhibition of tool wear. A method is needed that can monitor the accumulation state in real time and comprehensively, and actively regulate and control according to the spatio-temporal evolution characteristics of the accumulation, in order to improve the processing quality and efficiency, and prolong the service life of the tool and equipment. There is a lack of an accumulation state active control and tool wear inhibition method for the tool barrel based on multi-sensor fusion and active control model in the prior art, which cannot accurately predict the spatio-temporal evolution characteristics of the accumulation deposition and dynamically regulate and control the cooling medium, and is difficult to meet the needs of modern mechanical processing. Therefore, it is urgent to develop a new method to solve these problems. SUMMARY

[0006] The purpose of the present application is to provide an accumulation state active control and tool wear inhibition method for the tool barrel to solve the problems raised in the background art.

[0007] To achieve the above-mentioned purpose, the present application provides the following technical solution: an accumulation state active control and tool wear inhibition method for the tool barrel applied to a mechanical processing control system, the system comprising a plurality of accumulation monitoring sensors arranged on the inner wall of the tool barrel and an active control module connected to the accumulation monitoring sensors, two adjacent accumulation monitoring sensors being apart by a set distance, the method comprising: acquiring vibration frequency spectrum data and temperature gradient data of the inner wall of the tool barrel through the accumulation monitoring sensors to generate an accumulation state data set; receiving real-time accumulation state data of the plurality of accumulation monitoring sensors through the active control module to construct an accumulation state matrix of the tool barrel; determining spatio-temporal evolution characteristics of the accumulation deposition according to the accumulation state data set and the real-time data of each accumulation monitoring sensor, wherein determining the spatio-temporal evolution characteristics of the accumulation deposition comprises processing the accumulation state matrix, extracting deposition characteristics in combination with the accumulation state data set, predicting the accumulation trend according to the temperature gradient change and the pressure distribution information, outputting the spatio-temporal evolution characteristics of the accumulation deposition through an active control model, and updating the accumulation state data set according to the spatio-temporal evolution characteristics; dynamically regulating and controlling the cooling medium of the tool barrel according to the spatio-temporal evolution characteristics.

[0008] Preferably, the determination of the spatio-temporal evolution characteristics of the accumulation deposition comprises: processing the accumulation state matrix to extract vibration frequency spectrum characteristics, temperature distribution characteristics and accumulation growth trend; modeling the accumulation state matrix according to the vibration frequency spectrum characteristics and the temperature distribution characteristics, dividing the inner wall of the tool barrel into a plurality of sub-sections and marking section identifiers, associating and matching the vibration frequency spectrum of the sub-sections with the accumulation state data set, and marking the section identifiers in the accumulation state data set; According to the temperature gradient change calculated according to the positions of the slag monitoring sensors, the evolution of the slag deposition is predicted according to the temperature gradient change and the slag growth trend, and deposition prediction information of each sub-section is calculated; An active control model is constructed, the deposition prediction information is taken as an input parameter of the active control model, the deposition prediction information is spatially correlated and modeled by the active control model, and a spatio-temporal evolution feature of the slag deposition is output; The slag state dataset is updated according to the spatio-temporal evolution feature, and the spatio-temporal evolution feature of the slag deposition is obtained.

[0009] Preferably, the processing of the slag state matrix comprises: The slag state matrix is standardized, a deposition hot spot section in the matrix is intercepted by a sliding window, noise filtering is performed on the hot spot section, and a slag growth trend is calculated by a feature decomposition algorithm; The spatial correlation feature of the slag state matrix is calculated, the adhesion strength, deposition stability coefficient and slag blind area index between sections are calculated according to the spatial correlation feature, a feature fusion network is constructed, and the temperature distribution characteristics are calculated by the feature fusion network; The time domain feature and frequency domain feature collected by each slag monitoring sensor are extracted, the deposition feature vector of the sensor is calculated according to the phase difference between the time domain feature and the frequency domain feature, the slag monitoring sensors at different positions are matched according to the deposition feature vector, and the slag growth trend is calculated.

[0010] Preferably, the deposition strength modeling of the slag state matrix comprises: According to the vibration spectrum feature, a vibration spectrum sampling point is extracted in each frame of data, the sampling point is associated and mapped with the temperature distribution characteristics, a deposition strength atlas is generated, the deposition strength atlases collected by multiple sensors are spatially aligned, and the deposition strength distribution of the cutter barrel is calculated; An adhesion threshold value is set, an adhesion source is located according to the vibration spectrum values of multiple frames of slag state matrices, an adhesion strength difference value is calculated, if the adhesion strength difference value is greater than or equal to the adhesion threshold value, it indicates that there is a slag blind area in the section, a flow field model constraint compensation is performed on the current section, the deposition strength distribution of the current section is iteratively corrected according to the fluid dynamics model corresponding to the current section, and a blind area deposition strength compensation value is calculated according to the correction result; The deposition strength distribution is used to model the deposition strength of the slag state matrix, and the deposition model of the cutter barrel is temperature-labeled by the temperature distribution characteristics.

[0011] Preferably, the calculation of the temperature gradient change according to the positions of the slag monitoring sensors comprises: According to a plurality of groups of accumulated slag deposition data, a temperature change point is extracted, and the change point is mapped to a unified coordinate system according to the deployment position of the sensor, the change point is fitted through a spatial interpolation algorithm, and an accumulated slag field model of the cutter barrel is generated; Equal-interval sampling is performed along a flow path of the accumulated slag field model, a deposition change parameter is calculated according to a sampling result, a deposition change parameter is calculated according to the adhesion coefficient change rate, the temperature fluctuation index and the deposition change slope of the path; According to the deposition monitoring sensor deployment parameters and the collection accuracy, the accumulated slag growth in the space-time evolution characteristics of each frame of data is projected to the accumulated slag field model, the accumulated slag growth evolution law in the partition is analyzed according to the number of sensors along the flow direction of the accumulated slag field model, and the deposition distribution characteristics are calculated according to the evolution law. According to the deposition change parameter and the deposition distribution characteristics, a temperature gradient change is calculated, and the calculation process of the temperature gradient change includes: selecting a spatial coordinate point in the sensor deployment direction based on the position range from the first accumulated slag monitoring sensor to the last accumulated slag monitoring sensor, accumulating and calculating the product of the deposition field intensity characteristic weight value and the accumulated slag distribution characteristic weight value in the spatial resolution range, and superimposing the influence value of the sensor collection frequency on the deposition intensity change rate.

[0012] Preferably, the calculation of the deposition prediction information of each sub-section includes: Taking the flow main path of the accumulated slag field model as a reference line and taking the peak position of the accumulated slag growth in each frame of data as a reference point, a deposition offset is calculated, and a deposition distribution curve is drawn according to coordinates; According to the temperature gradient change, the growth rate and direction in the accumulated slag growth trend are corrected; Starting from the nearest deposition distribution point, the distribution curve is continuously drawn according to the correction results of the growth rate and direction to generate the next time period deposition distribution point, until the distribution point covers the entire target section to generate the deposition prediction information.

[0013] Preferably, the active control model includes: An input layer is used to organize the deposition prediction information into spatial distribution data and perform standardization processing; A feature fusion layer is used to extract the section correlation characteristics of the deposition by processing the spatial distribution data, and to construct the dependency relationship between physical units; A medium regulation layer is used to integrate the correlation relationship of the accumulated slag deposition on the space unit, and to generate a cooling medium regulation strategy.

[0014] Preferably, the space-time evolution characteristics of the accumulated slag deposition include: According to the space-time evolution characteristics of the accumulated slag deposition output by the active control model, the identification of the sub-section is corresponded to the space-time evolution characteristics. The section data in the accumulated slag state data set is reorganized according to the space-time characteristics to generate a section distribution diagram sorted by the deposition intensity of the accumulated slag; According to the reorganized section distribution diagram, the optimized deposition evolution characteristics of the accumulated slag are output.

[0015] Preferably, the dynamic regulation of the cutting barrel cooling medium comprises: The section identifier is one-to-one mapped with the section of the space-time evolution characteristics of the accumulated slag deposition; According to the space-time evolution characteristics of the accumulated slag deposition, the execution action of the cooling medium is controlled, including flow pressure adjustment, cooling liquid ratio and spray angle operation; According to the spatial distribution of the accumulated slag deposition and the preset medium regulation strategy, the cooling medium is dynamically allocated to the corresponding physical section.

[0016] Preferably, the execution action of the cooling medium comprises: When the accumulated slag deposition in the target section reaches the preset intensity threshold, the flow pressure increase instruction of the adjacent nozzle is triggered; According to the cooling liquid ratio strategy, the additive components are dynamically combined to generate a spray angle vector; The pressure parameter of the target section spray unit is adjusted based on the spray angle vector.

[0017] Compared with the prior art, the beneficial effects of the present application are: In the aspect of accumulated slag state monitoring, the vibration frequency spectrum data and temperature gradient data are collected by the plurality of accumulated slag monitoring sensors arranged on the inner wall of the cutting barrel to generate an accumulated slag state data set, which can comprehensively and real-timely obtain the accumulated slag related information of the inner wall of the cutting barrel. The layout of the plurality of sensors and the distance between adjacent sensors are set to ensure the spatial coverage of the accumulated slag state of the inner wall of the cutting barrel, avoid the limitation of single sensor monitoring, and enable the distribution and change of the accumulated slag at different positions to be captured. The vibration frequency spectrum data can reflect the change of the cutting barrel vibration characteristics caused by the accumulated slag deposition, and the temperature gradient data can reflect the heat conduction characteristics in the accumulated slag deposition process, and the combination of the two provides a rich information basis for accurately judging the accumulated slag state.

[0018] In the aspect of analysis and modeling of the accumulation state, the active control module receives real-time data from multiple sensors to construct an accumulation state matrix of the tool cylinder, and performs a series of processing on it, including standardization processing, noise filtering, feature decomposition, etc., which can extract key information such as vibration frequency spectrum characteristics, temperature distribution characteristics and accumulation growth trend. By dividing the inner wall of the tool cylinder into multiple sub-sections and identifying the sections, a fine analysis of the accumulation state is realized. The spatial correlation feature calculation of the accumulation state matrix can analyze the adhesion strength between sections, the deposition stability coefficient and the accumulation blind area index, etc., and construct a feature fusion network to calculate the temperature distribution characteristics, making the analysis of the accumulation state more comprehensive and in-depth. In the process of deposition intensity modeling, the deposition intensity map is generated by the correlation mapping of vibration frequency spectrum characteristics and temperature distribution characteristics, combined with the adhesion threshold value for adhesion source positioning and flow field model constraint compensation, which can accurately calculate the deposition intensity distribution and the strength compensation value of blind area deposition, improving the accuracy and reliability of the accumulation state modeling.

[0019] In the aspect of accumulation deposition evolution prediction, the temperature gradient change is calculated according to the position of the accumulation monitoring sensor, the accumulation field model is generated by spatial interpolation algorithm fitting, and the deposition change parameters are calculated by equal interval sampling along the flow path, combined with the deposition distribution characteristics and the temperature gradient change, which can comprehensively consider the changes of various physical parameters in the accumulation deposition process. Taking the main flow path of the accumulation field model as the reference line, the growth rate and direction of the accumulation growth trend are corrected combined with the temperature gradient change, to generate deposition prediction information, realize the distribution prediction of the accumulation deposition in the next period, and provide a reliable basis for active control.

[0020] In the aspect of active control model construction and application, the active control model includes input layer, feature fusion layer and medium regulation layer, which can process the spatial distribution data of deposition prediction information, extract section correlation features, construct the dependence relationship between physical units, and then generate cooling medium regulation strategy. Through this model, intelligent conversion from accumulation state data to cooling medium regulation strategy is realized, so that the regulation of cooling medium can be dynamically adjusted according to the spatio-temporal evolution characteristics of accumulation deposition, improving the pertinence and effectiveness of regulation.

[0021] In the aspect of dynamic regulation of cooling medium, the section identification and spatio-temporal evolution characteristics of accumulation deposition are one-to-one mapped, the flow pressure adjustment, cooling liquid ratio and jet angle operation of the cooling medium are controlled according to the spatio-temporal evolution characteristics, and the cooling medium is dynamically allocated to the corresponding physical section. When the accumulation deposition in the target section reaches the preset intensity threshold, the flow pressure instruction of the adjacent nozzle is triggered, the additive composition is dynamically combined according to the cooling liquid ratio strategy, and the pressure parameter of the target section jet unit is adjusted based on the jet angle vector, realizing the precise control of the cooling medium. This dynamic regulation strategy can effectively inhibit the formation and development of accumulation, and reduce the wear of the tool caused by accumulation. Attached Figure Description

[0022] Figure 1 This is a schematic diagram illustrating the working principle of the active control method for slag accumulation in the cutter barrel and the method for suppressing cutter wear described in this invention. Figure 2 This is a schematic diagram illustrating the working principle of the slag state matrix processing method. Figure 3 This is a diagram illustrating the working principle of the temperature gradient change calculation method. Figure 4 This is a schematic diagram illustrating the working principle of the active control model. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figures 1-4 The present invention relates to a method for active control of slag accumulation in a tool barrel and suppression of tool wear, which is applied to a machining control system comprising multiple slag accumulation monitoring sensors disposed on the inner wall of the tool barrel and an active control module connected to the slag accumulation monitoring sensors, wherein adjacent slag accumulation monitoring sensors are spaced apart by a set distance, and specifically includes the following steps: The slag accumulation monitoring sensors collect vibration spectrum data and temperature gradient data of the inner wall of the cutter barrel to generate a slag accumulation status dataset. The slag accumulation monitoring sensors are deployed at predetermined intervals on the inner wall of the cutter barrel, covering key areas prone to slag accumulation, and synchronously collect vibration and temperature data in real time, forming a raw dataset containing spatiotemporal information.

[0025] The active control module receives real-time slag accumulation status data from multiple slag accumulation monitoring sensors and constructs a slag accumulation status matrix for the cutter barrel. Based on the sensor position coordinates and acquisition time sequence, the active control module maps multidimensional data to a matrix structure, with the matrix rows and columns corresponding to spatial location and time sequence, respectively, thus achieving a structured representation of the slag accumulation status.

[0026] Based on the slag state dataset and real-time data from each slag monitoring sensor, the spatiotemporal evolution characteristics of slag deposition are determined. The specific process includes: processing the slag state matrix, extracting depositional features from the slag state dataset, predicting slag trends based on temperature gradient changes and pressure distribution information, outputting the spatiotemporal evolution characteristics of slag deposition through an active control model, and updating the slag state dataset based on these spatiotemporal evolution characteristics.

[0027] According to the spatio-temporal evolution characteristics, the cooling medium of the tool cylinder is dynamically regulated. The active control module generates medium regulation instructions based on the spatio-temporal evolution characteristics, and drives the cooling system to adjust parameters such as flow, pressure, ratio and injection angle, so as to realize active intervention on the deposition of accumulated slag.

[0028] Embodiment 1: The determination of the spatio-temporal evolution characteristics of the accumulated slag deposition specifically includes processing of the accumulated slag state matrix, deposition intensity modeling, temperature gradient change calculation, deposition prediction information generation and active control model construction, etc. Each link is related to each other to form a complete accumulated slag evolution characteristic analysis process.

[0029] The processing of the accumulated slag state matrix is the basis for determining the evolution characteristics. The accumulated slag state matrix is constructed by the active control module according to the real-time data of multiple accumulated slag monitoring sensors. The rows and columns of the matrix correspond to the spatial positions and data collection time sequences of the sensors, respectively. In the processing process, first, the matrix is standardized to eliminate the dimensional differences of the data, so that the vibration frequency spectrum data and the temperature gradient data of different sensors are comparable. After standardization, the sliding window technology is used to intercept the deposition hot spot section in the matrix. The size and moving step of the window are dynamically adjusted according to the processing conditions of the tool cylinder and the historical data of the accumulated slag, so as to ensure that the area of rapid change of the accumulated slag can be accurately captured. Noise filtering is performed on the hot spot section. The median filter or Gaussian filter algorithm can be used to remove high-frequency noise and random interference and retain the signal characteristics reflecting the essence of the accumulated slag deposition. Subsequently, the processed matrix is reduced in dimension through a feature decomposition algorithm (such as principal component analysis PCA), and the principal component parameters that can represent the growth trend of the accumulated slag are extracted. These parameters include the accumulated slag thickness growth rate, the deposition area expansion speed, etc., so as to realize the quantitative description of the growth trend of the accumulated slag.

[0030] In the deposition intensity modeling link, the accumulated slag state matrix needs to be analyzed in combination with the vibration frequency spectrum characteristics and the temperature distribution characteristics. First, the inner wall of the tool cylinder is divided into multiple sub-sections. The division of the sub-sections is based on the deployment position of the sensors and the structural characteristics of the tool cylinder, for example, the tool cylinder can be divided equidistantly along the axial direction, and each sub-section corresponds to a unique section identifier. Through the vibration frequency spectrum feature extraction algorithm, the feature frequency components related to the deposition of the accumulated slag are identified from each frame of data. For example, when the thickness of the accumulated slag increases, the vibration energy of a specific frequency will change significantly. These sampling points corresponding to the feature frequencies are extracted. The sampling points are associated and mapped with the temperature distribution characteristics, for example, high temperature areas usually correspond to positions where the accumulated slag is easy to adhere. By establishing a two-dimensional mapping relationship between vibration energy and temperature, a deposition intensity map is generated. The deposition intensity maps generated by multiple sensors are spatially aligned, that is, the local maps are spliced into a global deposition intensity distribution map according to the physical coordinates of the sensors, so as to intuitively reflect the accumulated slag intensity difference of each region of the inner wall of the tool cylinder.

[0031] In the deposition intensity modeling process, an adhesion threshold value is set to identify the blind zone of the deposit. The adhesion threshold value is determined by analyzing historical deposit data, which is usually a relative change threshold of the vibration spectrum value. When the vibration spectrum value of multiple frames of deposit state matrix changes beyond the threshold value, it indicates that there may be a deposit blind zone in the corresponding section that the sensor cannot directly monitor. At this time, the flow field model is used to constrain and compensate the deposition intensity of the section. The flow field model is based on the principle of fluid dynamics, and simulates the deposition and scouring process of the deposit under the action of the fluid by calculating the flow velocity and pressure distribution of the cooling liquid in the section. Through iterative correction of the measured data and the predicted deposition intensity, the intensity compensation value of the blind zone deposition is calculated to correct the estimation error of the deposition intensity distribution. Finally, combined with the deposition intensity distribution and temperature distribution characteristics, the deposit state matrix is labeled to form a deposition model containing thermal-mechanical coupling characteristics.

[0032] The calculation of temperature gradient change is a key link in deposit trend prediction. First, according to multiple sets of deposit deposition data, points with significant temperature changes (i.e. positions where the temperature gradient exceeds a set threshold) are extracted and mapped to a unified coordinate system, which is established as a cylindrical coordinate system with the cutter barrel axis as the reference. Through a spatial interpolation algorithm (such as Kriging interpolation or polynomial interpolation), the temperature change points are fitted to generate a continuous deposit field model, which reflects the corresponding relationship between the temperature field of the inner wall of the cutter barrel and the deposit distribution. Along the flow path (i.e. the direction of cooling liquid flow) of the deposit field model, equidistant sampling is performed, and the spacing of the sampling points is determined according to the model resolution and the evolution speed of the deposit, usually 1 / 2 to 1 times the spacing of the sensor. The adhesion coefficient change rate, temperature fluctuation index and deposition change slope are calculated for each sampling point, where the adhesion coefficient change rate reflects the change in the scouring ability of the cooling liquid to the deposit, the temperature fluctuation index represents the stability of the local thermal environment, and the deposition change slope is the rate of change of the deposit thickness with time. The three parameters are summed by weighting to generate a deposition change parameter, and the weight coefficients are determined according to the historical experience of the processing technology.

[0033] In addition, according to the deployment parameters (such as spacing, angle) and the collection accuracy of the sensor, the deposition growth evolution characteristics in each frame of data are projected into the deposition field model, and the model is divided into multiple analysis partitions along the flow direction, and the number of partitions matches the number of sensors. By statistically analyzing the mean, variance and other characteristics of the deposition growth in each partition, the deposition distribution characteristics are calculated, such as the spatial distribution range of the high deposition rate area. Finally, the calculation of the temperature gradient change comprehensively considers the deposition change parameters and the deposition distribution characteristics. The specific process is as follows: a series of spatial coordinate points are selected in the direction of sensor deployment. For each coordinate point, the product of the deposition field intensity characteristic weight value and the deposition distribution characteristic weight value in its spatial resolution range is calculated, and the weight value decays exponentially with the increase of spatial distance. At the same time, the influence value of the sensor collection frequency on the deposition intensity change rate is superimposed, which is calculated through the linear relationship model of the collection frequency and the deposition rate, thereby forming a complete temperature gradient change parameter.

[0034] In the deposition prediction information generation link, the flow main path (i.e. the direction of the main flow of the cooling liquid) of the deposition field model is taken as the reference line, and the peak position of the deposition growth in each frame of data is taken as the reference point. The deposition offset of the reference point relative to the reference line is calculated, which reflects the spatial distribution deviation of the deposition of the deposition. According to the offset and the coordinate data, the deposition distribution curve is drawn, and the horizontal coordinate is the spatial position and the vertical coordinate is the deposition intensity. The growth rate and direction in the deposition growth trend are corrected using the temperature gradient change parameter, for example, in the high temperature area, the deposition growth rate may accelerate, and the parameters of the prediction model need to be adjusted accordingly. Starting from the latest deposition distribution point, the distribution curve is drawn according to the corrected growth rate and direction, and the deposition distribution point of the next period is generated point by point until the entire target section is covered, forming a complete deposition prediction information, which contains the deposition intensity prediction value and the spatial distribution form of each sub-section in the future period.

[0035] The construction of the active control model is the core of realizing the analysis of the evolution characteristics of the accumulated slag. The model includes an input layer, a feature fusion layer, and a medium regulation layer. The input layer receives the deposition prediction information and organizes it into a spatial distribution data matrix, with each element of the matrix corresponding to a predicted parameter (such as accumulated slag intensity, growth rate) of a sub-section. At the same time, the data is standardized to meet the input requirements of the model. The feature fusion layer uses a graph neural network (GNN) or a convolutional neural network (CNN) structure to process spatial distribution data and extract associated features between sub-sections, such as the degree of mutual influence of adjacent section accumulated slag deposition, to build a dependency model between physical units. The output of the feature fusion layer is a high-dimensional vector containing spatial correlation features, which represents the overall evolution trend of the accumulated slag deposition on the inner wall of the cutter barrel. The medium regulation layer integrates the correlation of each spatial unit based on the output of the feature fusion layer to generate cooling medium regulation strategies, such as increased cooling fluid flow instructions or jet angle adjustment schemes for high deposition risk sections.

[0036] Finally, according to the spatio-temporal evolution characteristics output by the active control model, the sub-section identification is matched one-to-one with the evolution characteristics (such as deposition intensity, spatial distribution), and the section data in the accumulated slag state data set is reorganized according to the spatio-temporal characteristics to generate a section distribution map sorted by deposition intensity. By analyzing the reorganized distribution map, the optimized accumulated slag deposition evolution characteristics are extracted, which will serve as the basis for dynamic regulation of the cooling medium to achieve active control of the accumulated slag state in the cutter barrel.

[0037] Example 2: The processing of the accumulated slag state matrix includes key steps such as standardization, hot section extraction, noise filtering, feature decomposition, spatial correlation analysis, feature fusion network construction, and multi-sensor feature matching.

[0038] The standardization of the accumulated slag state matrix is the core step of data preprocessing. After receiving real-time data from multiple accumulated slag monitoring sensors, the active control module constructs a two-dimensional matrix containing vibration spectrum data and temperature gradient data. The rows of the matrix correspond to the spatial positions of the sensors, and the columns correspond to the time series. Since the measurement data of different types of sensors have different dimensions (such as frequency values of vibration spectrum and Celsius degrees of temperature), the matrix needs to be standardized to eliminate the influence of dimensional differences on subsequent analysis. The standardization method uses Z-score standardization, which calculates the difference between each data point and the mean value and divides it by the standard deviation, so that the processed data conforms to the standard normal distribution with a mean of 0 and a standard deviation of 1. Through standardization, the data of different physical quantities is comparable, facilitating subsequent feature extraction and model analysis.

[0039] After the standardization, the sliding window technique is used to intercept the deposition hotspot section in the matrix. The size of the sliding window is determined according to the time span of the rapid growth of the historical deposition data, for example, set to 10-20 sampling periods, and the window moving step is 1-5 sampling periods, to ensure the continuity and integrity of the hotspot section. The determination of the deposition hotspot section is based on the mutation of vibration spectrum energy or the significant change of temperature gradient. When the average energy of the data in the window exceeds the preset threshold (such as 1.5 times of the historical average), the window is determined as a hotspot section. The intercepted hotspot section is filtered for noise, and the median filtering algorithm is used to remove impulse noise, or the Gaussian filtering algorithm is used to smooth the data curve, and the true signal characteristics reflecting the deposition of the deposition are retained. After noise filtering, the matrix of the hotspot section is processed by dimension reduction through a feature decomposition algorithm (such as principal component analysis PCA), and the first k principal components (k is determined according to the cumulative variance contribution rate, usually the number of principal components with a cumulative contribution rate of more than 85%) are extracted. The coefficient vector of the principal component reflects the contribution degree of each original variable (vibration frequency, temperature, etc.) to the growth trend of the deposition, so as to realize the quantitative description of the growth trend of the deposition, such as the time series change of the principal component score can represent the growth rate of the deposition thickness.

[0040] The spatial correlation features of the deposition state matrix are calculated to analyze the deposition correlation between different sub-sections. The spatial correlation features are obtained by calculating the cross-correlation coefficient matrix of the data of each sub-section. The cross-correlation coefficient reflects the similarity of the data of two sections in the time sequence, and the larger the coefficient value (close to 1) indicates that the correlation of the deposition between the sections is stronger. Based on the cross-correlation coefficient matrix, the adhesion strength, deposition stability coefficient and deposition blind area index between the sections are further calculated. The adhesion strength is calculated by the peak frequency component energy of the cross-correlation coefficient of adjacent sections, which represents the possibility of the deposition adhesion between adjacent sections; the deposition stability coefficient is calculated by the ratio of the standard deviation to the mean value of the section data, which reflects the fluctuation degree of the deposition state of the section, and the smaller the coefficient indicates that the deposition is more stable; the deposition blind area index is calculated by the energy entropy of the section data, and the larger the entropy value indicates that the deposition state of the section is more complex, which may exist the situation of insufficient sensor coverage or monitoring blind area.

[0041] After obtaining the spatial correlation features, a feature fusion network is constructed to calculate the temperature distribution characteristics. The feature fusion network adopts a fully connected neural network structure, the input layer is the spatial correlation features (such as cross-correlation coefficient, adhesion strength, stability coefficient, etc.) and the temperature raw data, the hidden layer realizes the nonlinear transformation of the features through the nonlinear activation function (such as ReLU), and the output layer is the fused temperature distribution characteristic parameters, including the mean, variance, gradient, etc. of the temperature field. The feature fusion network is trained and optimized by historical data to ensure that the output temperature distribution characteristics can accurately reflect the coupling relationship between the deposition and the temperature field. For example, in the section where the deposition grows rapidly, the temperature distribution characteristics may present the characteristics of local high temperature and significant gradient change, and through the feature fusion network, these characteristics can be associated with the spatial correlation features, improving the accuracy of temperature field analysis.

[0042] In addition, for the data collected by each deposition monitoring sensor, its time domain features and frequency domain features are extracted. The time domain features include mean, variance, peak value, kurtosis, etc. statistical quantities, reflecting the overall characteristics of the vibration signal in the time dimension; the frequency domain features are obtained by converting the time domain signal to the frequency domain signal through fast Fourier transform (FFT), and extracting the frequency spectrum energy distribution, characteristic frequency (such as the inherent frequency of the deposition), etc. According to the phase difference between the time domain features and the frequency domain features, the deposition feature vector of the sensor is calculated. The phase difference analysis is used to identify the time delay relationship between different frequency components in the vibration signal and the deposition, for example, the high frequency vibration may lag behind the increase of the deposition thickness, and through the phase difference calculation, the time sequence correlation between the frequency components and the deposition state can be established. The deposition feature vector is composed of the normalized time domain features, frequency domain features and phase difference parameters, and each element of the vector corresponds to a feature dimension.

[0043] Based on the deposition feature vector, the feature matching is performed on the deposition monitoring sensors at different positions to identify the sensor group with similar deposition state. The feature matching adopts the Euclidean distance or cosine similarity measurement method to calculate the distance or similarity between the deposition feature vectors of any two sensors, the sensors with a distance less than a set threshold (such as 0.5) or a similarity greater than a set threshold (such as 0.8) are divided into the same group, indicating that the deposition states of the sections where these sensors are located are similar. Through feature matching, the sensors can be divided into several groups, each group corresponding to a typical deposition mode, thereby realizing the group analysis of the deposition growth trend. For example, the sensors in the same group may all be in the high flow cooling area of the cutter barrel, and the deposition growth trend in this area is greatly affected by the fluid scouring. Through group analysis, the deposition evolution law of this area can be extracted, providing a basis for subsequent zoning regulation.

[0044] On the basis of feature matching, the space distribution and data characteristics of the sensors in the group are combined to calculate the growth trend of the slag. For each group, the least squares method is used to fit the curve of the change of the slag thickness with time, and the slope of the curve is the growth rate of the slag corresponding to the group. At the same time, by analyzing the spatial position distribution of the sensors in the group, the main direction of the growth of the slag is determined, for example, the expansion trend along the axial or circumferential direction of the cutter barrel. In addition, considering the time synchronization of the data collected by the sensors, the growth rate and direction of the sensors in the group are weighted and averaged, and the weight is determined according to the collection accuracy of the sensors and the importance of the spatial position (such as the sensors near the cutting area of the cutter have higher weight). Finally, the growth trend parameters of the slag of the group are obtained, including the growth rate, the expansion direction, the spatial distribution range, etc.

[0045] In the whole process of slag state matrix processing, each link is closely connected: the standardization processing lays the data foundation for subsequent feature extraction, the sliding window and noise filtering ensure that the analysis focuses on the key area, the feature decomposition realizes data dimension reduction and trend quantization, the spatial correlation analysis reveals the deposition correlation between sections, the feature fusion network improves the accuracy of temperature field analysis, and the multi-sensor feature matching and group analysis realize the fine description of the growth trend of the slag in different regions.

[0046] Example 3: The deposition intensity modeling of the slag state matrix specifically includes the correlation mapping of vibration spectrum characteristics and temperature distribution characteristics, deposition intensity map generation, spatial alignment processing, adhesion threshold setting, flow field model constraint compensation, and temperature labeling, etc. Through multi-dimensional data fusion and physical model driving, the deposition intensity of the inner wall of the cutter barrel is accurately modeled and blind area compensation is realized.

[0047] The correlation mapping based on vibration spectrum feature extraction and temperature distribution characteristics is the basis of deposition intensity modeling. The vibration spectrum data collected by the slag monitoring sensor in real time contains characteristic frequency components reflecting the thickness and adhesion state of the slag, for example, when the deposition of the slag causes the local mass increase of the inner wall of the cutter barrel, the low-frequency component energy in the vibration spectrum will be significantly enhanced. In each frame of data, the sampling points of the vibration spectrum are extracted by peak detection algorithm, usually the top 5% of the energy frequency points are selected as the characteristic frequency, and the vibration energy values corresponding to these frequency points are directly related to the deposition intensity. At the same time, the temperature distribution characteristics reflect the differences in the thermal environment of different regions of the inner wall of the cutter barrel, and the high-temperature region may exacerbate the adhesion of the slag due to material thermal expansion, and the low-temperature region may cause the slag to fall off due to the flushing of the cooling liquid. By establishing a two-dimensional mapping relationship between the vibration spectrum sampling points and the temperature values, for example, using a polynomial fitting or a neural network model, the vibration energy values and the temperature values are converted into a deposition intensity index, which can be quantified as a dimensionless deposition intensity value (range 0-1), and the larger the value, the higher the deposition intensity of the slag.

[0048] A deposition intensity map is generated and spatially aligned. For each deposit monitoring sensor, the associated mapping of the deposition intensity indicator is arranged in time series according to its deployment position (such as the circumferential angle of the cutter barrel, the axial coordinate), forming a one-dimensional deposition intensity map corresponding to the sensor. The horizontal axis of the map is time, and the vertical axis is the deposition intensity value. To achieve global integration of multi-sensor data, the deposition intensity map of each sensor needs to be spatially aligned. The core of spatial alignment is to convert the local coordinate system of different sensors into the global cylindrical coordinate system of the cutter barrel (with the cutter barrel axis as the z-axis, the circumferential angle as the theta-axis, and the radial distance as the r-axis). Through the coordinate transformation matrix, the deposition intensity map of each sensor is mapped to the corresponding position in the global coordinate system, and then the bilinear interpolation algorithm is used to fill the blank area between adjacent sensors, and finally a two-dimensional deposition intensity distribution map covering the entire circumference of the inner wall of the cutter barrel is generated. The distribution map directly displays the spatial distribution characteristics of the deposition intensity of the deposits, for example, the deposition intensity near the cutting area of the cutter may be significantly higher than that in other areas.

[0049] In the deposition intensity modeling process, an adhesion threshold value is set to identify the deposit blind area. The adhesion threshold value is determined by analyzing the vibration frequency spectrum variation range of the historical deposition data, and is usually set to 2-3 times the standard deviation of the vibration energy fluctuation range under normal working conditions. When the difference between the vibration frequency spectrum values of multiple frames of deposit state matrix (i.e., the difference between the vibration energy of the current frame and the previous frame) exceeds the adhesion threshold value, it indicates that there may be a blind area where the deposits grow rapidly and are not directly monitored by the sensor. At this time, a beamforming algorithm is used to locate the adhesion source and determine the spatial position of the blind area. The beamforming algorithm estimates the azimuth angle of the signal source by calculating the phase difference of the signals of each sensor, thereby achieving preliminary positioning of the deposit blind area.

[0050] For the identified deposit blind area, flow field model constraint compensation is needed. The flow field model is based on the Navier-Stokes equation of fluid dynamics, combined with the geometric shape of the inner wall of the cutter barrel and the flow parameters of the cooling liquid (such as flow rate, pressure), and solves the fluid velocity field and pressure distribution of the section through the finite element method. The input parameters of the flow field model include the density, viscosity, inlet flow rate of the cooling liquid, and the rotational speed of the cutter barrel, etc. By comparing the predicted deposition intensity of the model with the measured value of the sensor, an iterative correction mechanism is established: if the error between the predicted value and the measured value exceeds the preset threshold (such as 10%), adjust the boundary conditions (such as wall roughness) or fluid parameters in the model, recalculate the deposition intensity distribution, until the error converges. Through iterative correction, the intensity compensation value of the blind area deposition is calculated, which is used to correct the estimated value of the blind area region in the deposition intensity distribution map, improving the modeling accuracy.

[0051] After the blind area compensation is completed, the deposition intensity modeling is performed on the deposition state matrix according to the deposition intensity distribution. The essence of the deposition intensity modeling is to replace each element (corresponding to a spatial point on the inner wall of the tool cylinder) in the deposition state matrix with a corresponding deposition intensity value to form a deposition intensity matrix. The matrix not only contains the current deposition intensity of each spatial point, but also retains the historical information of the evolution of the deposition through the time sequence. At the same time, the deposition model is temperature-labeled through the temperature distribution characteristics, that is, the temperature field information is superimposed in the deposition intensity matrix. Specifically, the deposition intensity value and the temperature value of each spatial point form a key-value pair and are stored in the matrix, so that the deposition model has the characteristics of thermal-mechanical coupling. For example, the deposition intensity value of a spatial point is 0.7 and the temperature value is 50°C, which is labeled as (0.7, 50°C) for subsequent analysis of the interaction between the deposition and the temperature field.

[0052] In the whole process of deposition intensity modeling, the correlation mapping of vibration spectrum characteristics and temperature distribution characteristics is the core of data-driven, and through the coupling analysis of characteristic frequency energy and temperature, a quantitative index of deposition is established; the spatial alignment processing realizes the global integration of multi-sensor data, solving the contradiction between local monitoring and global modeling; the adhesion threshold setting and flow field model constraint compensation are aimed at the deposition blind area problem, and the robustness of modeling is improved through physical model driving; the temperature labeling part injects the thermal environment dimension into the deposition model, making it closer to the actual processing conditions. Each step supports each other, forming a complete technical chain from data acquisition, feature extraction, spatial modeling to blind area compensation.

[0053] In addition, the results of the deposition intensity modeling can be further used for deposition trend prediction and cooling medium regulation. For example, by analyzing the time sequence change of the deposition intensity matrix, the expansion direction and rate of the deposition can be predicted; based on the deposition intensity distribution, the active control module can develop differentiated cooling strategies to increase the cooling liquid injection flow rate or adjust the injection angle in the high deposition intensity section, thereby achieving active inhibition of the deposition. At the same time, the temperature labeling information of the deposition intensity model can guide the adjustment of the cooling liquid ratio, for example, using cooling liquid additives with better heat dissipation performance in high temperature sections to reduce the possibility of deposition adhesion.

[0054] Embodiment 4: The calculation of the temperature gradient change according to the deposition monitoring sensor position includes steps such as temperature change point extraction, deposition field model construction, flow path sampling analysis, partition evolution rule analysis, and temperature gradient change comprehensive calculation. Through spatial interpolation, parameter calculation, and feature projection, the quantitative analysis of the correlation characteristics of the temperature field on the inner wall of the tool cylinder and the deposition is realized.

[0055] Temperature change points are extracted based on multiple sets of deposition data. The deposition monitoring sensors are deployed on the inner wall of the barrel at a set interval, and real-time temperature gradient data is collected. When the temperature change rate at a certain location exceeds the pre-set threshold (e.g., 0.5°C per minute), it is determined to be a temperature change point. These points reflect the local thermal state mutation caused by deposition, such as temperature jump in the adhesion area due to frictional heating. According to the deployment position of the sensor (axial coordinate z, circumferential angle θ), the temperature change points are mapped to a unified cylindrical coordinate system to form a discrete set of temperature feature points.

[0056] The temperature change points are fitted by a spatial interpolation algorithm to generate a deposition field model of the barrel. The interpolation algorithm uses the Kriging interpolation method, which is based on the theory of regionalized variables. By calculating the spatial autocorrelation function of the temperature change points, the temperature value at an unknown location is estimated. The Kriging interpolation formula is:

[0057] wherein, is the temperature value at the to-be-estimated location, is the temperature value of the th temperature change point, is the weight coefficient determined by the semi-variogram function, is the number of temperature change points participating in interpolation. Through this algorithm, a continuous temperature field distribution surface is generated, and the surface shape reflects the influence of deposition on heat conduction, such as the formation of a high temperature value area in a region with a large thickness of deposition.

[0058] Equal-interval sampling is performed along the flow path of the deposition field model. The flow path is defined as the main flow direction of the cooling liquid on the inner wall of the barrel (usually axial), and the sampling interval is set according to the model resolution and deposition evolution speed, for example, one sample point is collected every 10 mm. For each sampling point, three key parameters are calculated: adhesion coefficient change rate, temperature fluctuation index, and deposition change slope. The adhesion coefficient change rate reflects the change in the scouring ability of the cooling liquid on the deposition surface, which is obtained by calculating the ratio of the difference in adhesion coefficient between adjacent sampling points to the axial distance; the temperature fluctuation index is the ratio of the standard deviation to the mean value of the temperature value at the sampling point, which represents the stability of the thermal environment; the deposition change slope is obtained by fitting the curve of deposition intensity change over time, which reflects the growth rate of deposition.

[0059] According to the above three parameters, the deposition change parameter is calculated, and the formula is:

[0060] wherein, is the adhesion coefficient change rate, is the temperature fluctuation index, is the deposition change slope, The weight coefficient is determined by historical data training and satisfies The deposition change parameter comprehensively represents the dynamic characteristics of the deposition of the accumulated slag, and the greater the value, the more active the evolution of the accumulated slag.

[0061] According to the deployment parameters (such as spacing and angle) of the accumulated slag monitoring sensor and the collection accuracy, the accumulated slag growth evolution characteristics (such as deposition intensity and growth rate) in each frame of data are projected to the accumulated slag field model. The projection process is realized by coordinate transformation, which maps the sensor measurement values to the corresponding spatial positions of the model. Along the flow direction (axial direction) of the accumulated slag field model, the model is divided into multiple analysis partitions according to the number of sensors, and each partition corresponds to the monitoring range of one or more sensors. By statistical analysis of the mean, variance, skewness and other characteristics of the accumulated slag growth in the analysis partition, the deposition distribution characteristics are extracted, such as the axial position range of the high deposition rate partition and the spatial distribution form of the deposition intensity.

[0062] The calculation of the temperature gradient change integrates the deposition change parameter and the deposition distribution characteristic. The specific process is as follows: a series of spatial coordinate points are selected in the deployment direction (axial direction) of the sensor The product of the deposition field intensity characteristic weight value and the accumulated slag distribution characteristic weight value is calculated for each coordinate point, where is obtained by normalizing the deposition intensity value, and is calculated by the probability density function of the deposition distribution. The product results of all coordinate points are accumulated, and the influence value of the sensor collection frequency on the deposition intensity change rate is added (calculated by the linear relationship model between the collection frequency and the deposition rate, is the proportionality coefficient), and finally the temperature gradient change value is obtained, and the calculation formula is as follows:

[0063] The formula quantifies the comprehensive influence of spatial deposition characteristics and time sampling frequency on the temperature gradient, reflecting the spatiotemporal coupling effect of the deposition of the accumulated slag.

[0064] In the whole calculation process of the temperature gradient change, the accumulated slag field model serves as a data carrier, integrating the spatial distribution information of the temperature field and the deposition field; the flow path sampling and partition analysis realize the multi-scale decomposition of the dynamic characteristics of the accumulated slag evolution; and the introduction of the weight coefficient and the frequency influence value ensures the adaptability of the calculation result to the actual working conditions.

[0065] ​​Example 5: Dynamic regulation of the cutting barrel cooling medium includes section identification and evolution characteristic mapping, cooling medium execution action control, dynamic allocation strategy formulation, and key parameter adjustment, etc. Through space-time characteristic driving and physical unit linkage, precise intervention of cutting barrel slag deposition is realized.

[0066] Section identification and slag deposition evolution characteristic mapping is the basis of dynamic regulation. After determining the space-time evolution characteristics of slag deposition, the active control module divides the inner wall of the cutting barrel into sub-sections matching the sensor deployment density (such as dividing into a section every 20 mm along the axial direction and a sector every 30° in the circumferential direction), each sub-section is assigned a unique section identification (such as "Z1-θ1" representing the first section in the axial direction and the first sector in the circumferential direction). Through the characteristic matching algorithm, the evolution characteristics of each sub-section such as deposition intensity, growth trend, and temperature distribution are established in a one-to-one correspondence with the section identification, forming a "section identification-evolution characteristic" mapping table. The mapping table is updated in real time to ensure that the control instructions can accurately point to the slag risk area. For example, when the deposition intensity of section "Z3-θ2" is monitored to exceed the threshold value of 0.6 in 5 consecutive frames of data, the mapping table immediately marks this section as a high-risk area, triggering the cooling medium regulation process.

[0067] The execution action control of the cooling medium is the core link of dynamic regulation, which specifically includes flow pressure adjustment, cooling liquid ratio, and spray angle operation. Flow pressure adjustment is realized through proportional electromagnetic valves, and the active control module dynamically adjusts the cooling liquid flow according to the deposition intensity of the section: when the deposition intensity of a section is in the range of 0.3-0.6, the flow increases by 20%; when it exceeds 0.6, the flow increases by 50%, and the auxiliary nozzles of adjacent sections are started to cooperate with the flushing. The pressure adjustment range is set according to the rated parameters of the cooling system, for example, when the basic pressure is 0.5 MPa, the pressure of the high-risk section can be increased to 0.8-1.0 MPa to enhance the flushing force.

[0068] The cooling liquid proportioning strategy is based on a pre-set multi-component additive scheme. The additives include lubricating type (such as fatty acid ester), heat dissipation type (such as ethylene glycol), and anti-adhesion type (such as polytetrafluoroethylene micro powder), and the proportioning ratio of each additive is determined by a neural network model trained by historical data. For example, when the zone temperature exceeds 60°C and the deposition intensity is greater than 0.5, the model automatically selects a proportioning scheme of 60% heat dissipation type additive and 30% anti-adhesion type additive, dynamically mixes the components through a multi-channel metering pump, and generates a cooling liquid that adapts to the current thermal-mechanical environment. The generation of the spray angle vector is based on the spatial coordinates of the zone and the deposition morphology, and the optimal spray direction of the nozzle is determined through geometric calculation: taking the center of the zone as the target, calculating the angle between the nozzle axis and the connecting line of the target, combining the deposition offset direction (such as a 15° clockwise offset along the circumference), generating a three-dimensional spray angle vector containing pitch and azimuth, and driving the servo motor to adjust the nozzle posture.

[0069] The dynamic allocation of cooling medium is based on the spatial distribution of accumulated deposition and the pre-set control strategy. The active control module first divides the deposition intensity distribution map into high (>0.6), medium (0.3-0.6), and low (<0.3) risk level areas, and formulates differentiated allocation strategies for different levels: high-risk areas are preferentially allocated 80% of the cooling liquid flow, and all components are enabled; medium-risk areas are allocated 15% of the flow, and basic cooling liquid (without additives) is used; low-risk areas are allocated 5% of the flow, and only basic cooling function is maintained. Dynamic allocation is achieved through a multi-way electromagnetic valve group, with each electromagnetic valve corresponding to a physical zone or nozzle group. The module sends switching instructions according to the real-time mapping table to ensure that the cooling liquid flows accurately to high-demand areas. For example, when the zone "Z5-θ4" is marked as high-risk, the corresponding electromagnetic valve V5-4 is immediately fully opened, while the electromagnetic valves of low-risk zones are closed, achieving dynamic reallocation of flow.

[0070] When controlling the cooling medium to perform actions, the pre-set intensity threshold is a key condition for triggering differentiated control. The threshold is determined by the wear resistance of the tool cylinder material and the processing process requirements, for example, when the deposition intensity reaches 70% of the critical adhesion strength of the material (usually set to 0.5-0.7), the flow pressure boost instruction of the adjacent nozzle is triggered. The pressure boost process uses closed-loop control: the pressure sensor feeds back the nozzle outlet pressure in real time, and if the deviation between the actual pressure and the target pressure (such as 1.0 MPa) exceeds 5%, the active control module automatically adjusts the proportional valve opening until the pressure stabilizes. For example, when the deposition intensity of the zone "Z2-θ3" reaches 0.6, the adjacent nozzle groups N2-3 and N2-4 simultaneously enter the pressure boost mode, and the pressure gradually increases from 0.5 MPa to 0.9 MPa until the deposition intensity drops below the threshold.

[0071] The dynamic combination of coolant ratio and the adjustment of spray angle are the key means to suppress the accumulation of slag. The selection and mixing ratio of additive components are driven by real-time temperature field and deposition intensity data. For example, in the section where the temperature gradient changes sharply (>2℃ / mm), the proportion of anti-adhesion additive is automatically increased to 40% to reduce the adhesion of slag to the surface of the tool cylinder. The adjustment frequency of the spray angle vector matches the evolution speed of the slag. When the deposition offset increases by 5mm, the spray angle is recalculated and the nozzle is driven to rotate to ensure that the spray direction always points to the front area of the growing slag. For example, if the deposition distribution curve of a certain section is offset by 10mm along the axial direction, the pitch angle of the spray angle is adjusted by 5° to cover the new deposition area.

[0072] During the whole dynamic regulation process, the real-time mapping of section identification and evolution characteristics ensures the spatial accuracy of control instructions. The multi-parameter coordinated adjustment of cooling medium flow, ratio, and angle realizes the multi-dimensional suppression of slag deposition. The preset threshold and closed-loop control mechanism guarantee the stability and reliability of the regulation process. Through the above steps, the active control module can dynamically generate and execute the optimal regulation strategy according to the spatio-temporal changes of slag deposition, effectively suppress the tool wear caused by excessive deposition of slag, and improve the continuity and machining precision of the machining system.

[0073] It should be noted that, in this text, relational terms such as first and second are used merely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variant thereof are intended to cover non-exclusive inclusions, so that a process, method, article, or apparatus that includes a list of elements does not only include those elements, but also includes other elements not expressly listed, or inherent to such a process, method, article, or apparatus.

[0074] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements, and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for active control of built-up edge state and suppression of tool wear, applied to a machining control system, the system comprising a plurality of built-up edge monitoring sensors arranged on the inner wall of a tool cylinder and an active control module connected to the built-up edge monitoring sensors, two adjacent built-up edge monitoring sensors being separated by a set distance, characterized in that, The method comprises: Collecting vibration frequency spectrum data and temperature gradient data of the inner wall of the cutter barrel through the slag accumulation monitoring sensor to generate a slag accumulation state data set; Receiving real-time slag accumulation state data of multiple slag accumulation monitoring sensors through the active control module to construct a slag accumulation state matrix of the cutter barrel; According to the slag accumulation state data set and the real-time data of each slag accumulation monitoring sensor, determining the spatio-temporal evolution characteristics of slag deposition, wherein the determination of the spatio-temporal evolution characteristics of slag deposition comprises: processing the slag accumulation state matrix, extracting deposition characteristics in combination with the slag accumulation state data set, and predicting the slag deposition trend according to the temperature gradient change and the pressure distribution information, outputting the spatio-temporal evolution characteristics of slag deposition through the active control model, and updating the slag accumulation state data set according to the spatio-temporal evolution characteristics; According to the spatio-temporal evolution characteristics, dynamically regulating and controlling the cooling medium of the cutter barrel.

2. The method according to claim 1, wherein The determination of the spatio-temporal evolution characteristics of slag deposition comprises: Processing the slag accumulation state matrix, extracting vibration frequency spectrum characteristics, temperature distribution characteristics and slag growth trend; According to the vibration frequency spectrum characteristics and the temperature distribution characteristics, modeling the deposition intensity of the slag accumulation state matrix, dividing the inner wall of the cutter barrel into multiple sub-sections and marking section identifiers, associating and matching the vibration frequency spectrum of the sub-sections with the slag accumulation state data set, and marking the section identifiers in the slag accumulation state data set; According to the temperature gradient change calculated according to the positions of the slag accumulation monitoring sensors, predicting the evolution of slag deposition according to the temperature gradient change and the slag growth trend, and calculating the deposition prediction information of each sub-section; Constructing an active control model, taking the deposition prediction information as the input parameters of the active control model, spatially correlating and modeling the deposition prediction information through the active control model, and outputting the spatio-temporal evolution characteristics of slag deposition; According to the spatio-temporal evolution characteristics, updating the slag accumulation state data set to obtain the spatio-temporal evolution characteristics of slag deposition.

3. The method according to claim 2, wherein The processing of the slag accumulation state matrix comprises: Standardizing the slag accumulation state matrix, intercepting the deposition hot spot section in the matrix through a sliding window, filtering noise of the hot spot section, and calculating the slag growth trend through a feature decomposition algorithm; Calculating the spatial correlation characteristics of the slag accumulation state matrix, calculating the adhesion strength, deposition stability coefficient and slag blind area index between sections according to the spatial correlation characteristics, constructing a feature fusion network, and calculating the temperature distribution characteristics through the feature fusion network; Extracting the time domain features and frequency domain features collected by each slag accumulation monitoring sensor, calculating the deposition feature vector of the sensor according to the phase difference between the time domain features and the frequency domain features, performing feature matching on the slag accumulation monitoring sensors at different positions according to the deposition feature vector, and calculating the slag growth trend.

4. The method according to claim 3, wherein The deposition intensity modeling of the slag accumulation state matrix comprises: According to the vibration frequency spectrum characteristics, extracting vibration frequency spectrum sampling points in each frame of data, associating and mapping the sampling points with the temperature distribution characteristics to generate a deposition intensity spectrum, spatially aligning the deposition intensity spectra collected by multiple sensors, and calculating the deposition intensity distribution of the cutter barrel; An adhesion threshold is set, and the adhesion source is located according to the vibration frequency spectrum value of the multi-frame accumulated slag state matrix, the adhesion strength difference value is calculated, and if the adhesion strength difference value is greater than or equal to the adhesion threshold, it indicates that there is an accumulated slag blind area in the section, the flow field model constraint compensation is performed on the current section, the deposition intensity distribution of the current section is iteratively corrected according to the corresponding fluid dynamics model of the current section, and the intensity compensation value of the blind area deposition is calculated according to the correction result; According to the deposition intensity distribution, the deposition intensity of the accumulated slag state matrix is modeled, and the deposition model of the nozzle is temperature-labeled through the temperature distribution characteristics.

5. The method according to claim 4, wherein The temperature gradient change is calculated according to the positions of the accumulated slag monitoring sensors, including: According to a plurality of sets of accumulated slag deposition data, a temperature change point is extracted, and the change point is mapped to a unified coordinate system according to the deployment position of the sensor, the change point is fitted through a spatial interpolation algorithm, and an accumulated slag field model of the nozzle is generated; The flow path along the accumulated slag field model is sampled at equal intervals, the adhesion coefficient change rate, the temperature fluctuation index and the deposition change slope of the path are calculated according to the sampling results, and the deposition change parameter is calculated according to the adhesion coefficient change rate, the temperature fluctuation index and the deposition change slope; According to the deployment parameters and the collection accuracy of the accumulated slag monitoring sensors, the time-space evolution characteristics of the accumulated slag growth in each frame of data are projected to the accumulated slag field model, the accumulated slag field model is partitioned in the flow direction according to the number of sensors, the evolution law of the accumulated slag growth in the partition is analyzed, and the deposition distribution characteristics are calculated according to the evolution law. According to the deposition change parameter and the deposition distribution characteristics, the temperature gradient change is calculated, and the calculation process of the temperature gradient change includes: based on the position range from the first accumulated slag monitoring sensor to the last accumulated slag monitoring sensor, selecting a spatial coordinate point in the direction of sensor deployment, accumulating the product of the deposition field intensity characteristic weight value and the accumulated slag distribution characteristic weight value in the spatial resolution range, and superimposing the influence value of the sensor collection frequency on the deposition intensity change rate.

6. The method of claim 5, wherein the method is characterized by: The deposition prediction information of each sub-section is calculated, including: Taking the flow main diameter of the accumulated slag field model as a reference line and taking the peak position of the accumulated slag growth in each frame of data as a reference point, the deposition offset is calculated, and the deposition distribution curve is drawn according to the coordinates; According to the temperature gradient change, the growth rate and direction in the growth trend of the accumulated slag are corrected; Starting from the nearest deposition distribution point, the distribution curve is continued to be drawn according to the correction results of the growth rate and direction to generate the next time period deposition distribution point, until the distribution point covers the entire target section to generate the deposition prediction information.

7. The method of claim 2, wherein the method further comprises: determining whether the tool is in a state of accumulation of a residue; and controlling the tool in the state of accumulation of the residue. The active control model includes: An input layer for organizing the deposition prediction information into spatial distribution data and performing standardization processing; A feature fusion layer for extracting the section correlation characteristics of the deposition by processing the spatial distribution data and constructing the dependency relationship between physical units; A medium regulation layer for integrating the correlation relationship of the accumulated slag deposition on the spatial unit to generate a cooling medium regulation strategy.

8. The method of claim 2, wherein the method is characterized by: The time-space evolution characteristics of the accumulated slag deposition are obtained, including: According to the time-space evolution characteristics of the accumulated slag deposition output by the active control model, the identification of the sub-section is corresponded to the time-space evolution characteristics. The section data in the accumulated slag state data set is reorganized according to the space-time characteristics to generate a section distribution diagram sorted according to the accumulated slag deposition intensity; According to the reorganized section distribution diagram, an optimized accumulated slag deposition evolution characteristic is output.

9. The method of claim 1, wherein the method further comprises: determining a state of the tool holder; and determining a state of the tool based on the determined state of the tool holder. The dynamic regulation of the cooling medium of the tool barrel comprises: The section identifier is one-to-one mapped with the section of the space-time evolution characteristic of the accumulated slag deposition; According to the space-time evolution characteristic of the accumulated slag deposition, the execution action of the cooling medium is controlled, including flow pressure adjustment, cooling liquid proportioning and jet angle operation; According to the spatial distribution of the accumulated slag deposition and the preset medium regulation strategy, the cooling medium is dynamically distributed to the corresponding physical section.

10. The method of claim 9, wherein the method is characterized by: The execution action of the cooling medium comprises: When the accumulated slag deposition in the target section reaches a preset intensity threshold, a flow pressure increase instruction of the adjacent nozzle is triggered; According to the cooling liquid proportioning strategy, the additive components are dynamically combined to generate a jet angle vector; Based on the jet angle vector, the pressure parameter of the jet unit of the target section is adjusted.

Images