Gun barrel deep hole processing method and device based on magnetostrictive actuator

Abstract

This invention relates to the field of deep machining of cutting inserts, and provides a method and apparatus for deep hole machining of artillery barrels based on a magnetostrictive actuator. The machining method includes installing a multi-source data sensor network system on a boring bar, a cutting tool, and the artillery barrel; processing the real-time acquired multi-source signals to extract time-frequency information from acoustic emission and vibration signals; establishing a tool wear prediction model based on the characteristics of deep hole machining of artillery barrels; inputting the time-frequency information into the tool wear prediction model to obtain the current signal of the control system; and inputting the current signal into a coil to control the extension and retraction of the magnetostrictive element, keeping the cutting tool and the artillery barrel in contact, thereby achieving fine adjustment of the tool displacement. The machining apparatus includes a spindle box, guide rails, a boring bar assembly, a cutting tool, and a floating support. This invention solves the problem of tool wear and time-varying excitation affecting the dynamic characteristics of boring bars with large length-to-diameter ratios, meeting the precision requirements of deep hole manufacturing of artillery barrels.

Description

Technical Field

[0001] This invention relates to the field of deep hole machining / cutting inserts, and particularly to a method and apparatus for deep hole machining of artillery barrels based on a magnetostrictive actuator. Background Technology

[0002] The barrel, as a crucial component of artillery, significantly impacts its firing accuracy. A gun barrel is defined as one with a length-to-diameter ratio greater than 20:1. In artillery barrel manufacturing, the inner diameter is typically 100-400 mm, and the total length is usually 7-12 m, with a length-to-diameter ratio reaching up to 120. As the initial stage of rifling machining, the dimensional accuracy, straightness, and roundness of deep-hole precision boring directly affect the quality of the rifling. However, with the application of high-strength and high-toughness gun steel, its high hardness and toughness cause rapid tool tip wear, and the nonlinear dynamics of the large length-to-diameter ratio boring bar cause time-varying tool tip trajectories. This makes it highly susceptible to exceeding machining tolerances during a single long boring stroke.

[0003] To address the aforementioned issues, boring tool radial micro-displacement control technology offers a novel approach to effectively compensate for tool tip trajectory deviations. While eccentric boring micro-compensation devices based on hydraulic or electromechanical driven radial micro-displacement mechanisms have been manufactured, hydraulic drives suffer from system complexity, and electromagnetically driven radial micro-displacement mechanisms are complex to design and difficult to achieve precise displacement control. Furthermore, displacement conversion devices are structurally complex, require extremely high rigidity and transmission accuracy, are prone to wear over long-term use, and struggle to meet high-frequency response speed requirements.

[0004] Building upon this foundation, with the development of micro-displacement deformation actuation technology using materials such as piezoelectric ceramics and giant magnetostrictive materials (GMMs), high-frequency response and high-precision micro-feed control of cutting tools have been achieved. Micro-displacement actuators driven by piezoelectric ceramics require very high driving voltages, making energy transmission difficult, and individual piezoelectric ceramic elements have low energy density, resulting in small output force and displacement. GMMs, on the other hand, offer advantages such as high energy density, large output force and displacement, and fast response speed, leading to the emergence of many high-performance giant magnetostrictive actuators (GMAs).

[0005] Therefore, this invention introduces the GMA micro-displacement technology of the giant magnetostrictive actuator into the deep hole boring system of the gun barrel. By controlling the micro-displacement of the boring tool with high frequency and high precision, the trajectory deviation of the tool tip is compensated, thereby achieving precision control of the gun barrel and improving the machining accuracy. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a method and apparatus for deep hole machining of artillery barrels based on a magnetostrictive actuator. Addressing the issues of cutting tool depth machining and excessive tolerances in artillery barrel machining, the invention first installs a multi-source data sensor network system at corresponding positions on the boring bar, cutting tool, and artillery barrel. Then, the time-frequency information from the multi-source signals is input into a tool wear prediction model. Through a high-frequency, high-precision boring tool micro-displacement instantaneous stabilization dual-control strategy embedded with the magnetostrictive actuator, online accurate prediction of tool tip wear under complex and variable working conditions is conducted. The time-varying dynamic characteristics of the boring bar based on time-varying cutting force and time-varying overhang are studied, and the time-varying law of the boring tool tip trajectory is analyzed. Finally, the extension and retraction of the magnetostrictive element are controlled, driving the push rod to perform micron-level displacement, keeping the cutting tool and artillery barrel in contact, thereby achieving fine-tuning of the tool displacement, effectively controlling machining errors, and realizing precise machining of deep holes in artillery barrels.

[0007] This invention provides a method for deep hole machining of artillery barrels based on magnetostrictive actuators, the specific implementation steps of which are as follows:

[0008] S1. Mount the gun barrel to be processed on the guide rail with a floating support, install the cutting tool on the working end of the boring bar, connect the drive end of the boring bar to the spindle box, and mount the fixed end of the boring bar on the guide rail with a floating support.

[0009] S2. Install the multi-source data sensor network system at the corresponding positions on the boring bar, cutting tool, and gun barrel, and set the cutting parameters of the cutting tool.

[0010] S3. Start the spindle box and the multi-source data sensor network system respectively, process the acoustic emission signal and vibration signal collected in real time by the multi-source data sensor network system, and extract the time-frequency information about the tool tip displacement from the acoustic emission signal and vibration signal.

[0011] S4. Based on the characteristics of deep hole machining of artillery barrels, establish a tool wear prediction model, which includes the following sub-steps:

[0012] S41. Using the state update expression of a bidirectional recurrent neural network, with the acoustic emission signal and vibration signal collected in step S3 as the core input and the tool tip displacement as the output, a tool tip wear prediction model under a single working condition is established. The nonlinear mapping relationship between the characteristics of multi-source signals and the tool tip wear-sensitive characteristics is explored. The state update expression of the bidirectional recurrent neural network is as follows:

[0013] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0014] i t =σ(Wi ·[h t-1 ,x t ]+b i )

[0015]

[0016] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0017] h t =o t *tanh(C t )

[0018] Among them, f t This indicates the forgetting of historical signal data during processing, i t C represents the acceptance of new signal data during processing. t This indicates the long-term storage and memory of signal data during the processing, which is updated at each time step. This indicates that the data generated based on the current input data is used to update C. t The information of the state, h t This indicates adjustments to the current forecast during the processing. t For the final data output, x t This represents the signal collected from the multi-source data sensor network system at the current moment, σ represents the Sigmoid activation function, ensuring that the signal output value is between (0,1), and W f W represents the weight matrix of the forget gate. i W represents the weight matrix of the input gate. C W represents the weight matrix for calculating the processing state update. o Let b represent the weight matrix of the output gate. f The bias term representing the forget gate, b i The input gate bias term, b C b represents the bias term used to update the current memory cell state. o The bias term h of the output gate t-1 C represents the hidden state at the previous moment. t-1 This indicates the stored state of the tool displacement compensation at the previous moment, and tanh represents the hyperbolic tangent activation function, which normalizes the signal data to the range (-1, 1).

[0019] S42. Based on the tool tip wear prediction model under a single working condition obtained in step S41, the single working condition source domain data sample set is used as input, and the target domain to be learned under the variable feed working condition is used as output. A feature extractor is constructed using a convolutional neural network and a bidirectional long short-term memory network to extract the multi-dimensional spatiotemporal features of the single working condition source domain data sample set and the target domain to be learned under the variable feed working condition, and a domain adaptive transfer model is constructed.

[0020] S5. Input the time-frequency information of the acoustic emission signal and vibration signal obtained in step S3 into the tool wear prediction model constructed in step S4, calculate the tool wear state, and generate the current signal of the control system.

[0021] S6. Input the current signal obtained in step S5 into the coil wound on the permanent magnet to change the magnetic field strength, thereby controlling the extension and retraction of the magnetostrictive element, driving the push rod to make a micron-level displacement, so that the tool tip and the cannon barrel are kept in contact, thereby realizing the fine adjustment of the tool displacement, continuously collecting and monitoring acoustic emission signals and vibration signals to form a closed-loop feedback control.

[0022] Preferably, the position distribution of the accelerometer on the boring bar is obtained by analyzing the sensitivity of the signal from the accelerometer at different installation positions on the boring bar to the wear of the tool tip using the Pearson correlation coefficient method. The expression for the Pearson correlation coefficient is as follows:

[0023]

[0024] Where, ρ X,Y These are the signals from the accelerometer at different mounting positions on the boring bar. X and Y are the correlation coefficients in the Pearson correlation coefficient method, cov(X,Y) are the covariances of X and Y, and σ is the signal from the accelerometer at different mounting positions on the boring bar. X and σ Y These are the standard deviations of X and Y, respectively.

[0025] Preferably, in step S2, the cutting parameters include cutting speed, feed rate, and current intensity; and in step S3, the multi-source signals include tool position deviation value, abnormal cutting force value, abnormal vibration value, and acoustic signal characteristic value.

[0026] Preferably, the specific implementation steps for step S3, which involves extracting time-frequency information from the acoustic emission signal and vibration signal, are as follows:

[0027] S31. Based on the characteristics of tool wear and machining conditions, a sliding frame is used to sample and extract signal data of a specified length. The dispersion of the signal data is analyzed, and outlier data points that exceed or fall below three times the standard deviation of the average value are removed to obtain the initial data sample.

[0028] S32. Based on the initial data samples, construct a single-condition source domain data sample set using the tool tip wear value as a label;

[0029] S33. Based on the single-condition source domain data sample set obtained in step S32, the low-frequency and high-frequency components of the single-condition source domain data sample set are simultaneously decomposed using wavelet transform to obtain frequency domain information. The expression for the wavelet transform method is:

[0030]

[0031] Among them, W a,b (t) represents the time-frequency local characteristics of the processing vibration, acceleration, and acoustic emission signals obtained after wavelet transform at different scales and locations, and x(t) represents the original signals acquired by multiple sensors during the actual processing. The complex conjugate of the mother wavelet function used in the wavelet transform is represented by , and a and b represent different levels and time characteristics of tool wear, respectively.

[0032] S34. The frequency domain information obtained in S33 is processed using the Hilbert-Huang method to extract the instantaneous frequency components of the vibration frequency change and cutting force fluctuation caused by tool wear, thus obtaining the time-frequency information in the multi-source signal. The expression for the Hilbert-Huang method is as follows:

[0033]

[0034] Where ω(t) represents the instantaneous frequency characteristics of the signal caused by the change in the contact characteristics between the tool and the gun barrel, and θ(t) represents the phase characteristics of the tool vibration and machining force as the wear degree changes.

[0035] Preferably, in step S31, the judgment expression for the abnormal data point is:

[0036]

[0037] Where z is the standardized value, X is the signal data, μ is the mean of the signal dataset, σ is the standard deviation of the signal dataset, and if |z|>3, the data is considered an outlier.

[0038] Preferably, in step S41, the expressions for the correlation analysis and kernel principal component analysis (KPCA) are as follows:

[0039]

[0040] Where φ(X) is the feature representation mapped to a high-dimensional space through the kernel function K, and α i It is the projection coefficient.

[0041] A second aspect of the present invention provides a deep hole machining apparatus for artillery barrels based on a magnetostrictive actuator, comprising a spindle box, a guide rail, a boring bar assembly, and a floating support. The fixed end of the spindle box is connected to a first mounting end of the guide rail, the output end of the spindle box is connected to a driving end of the boring bar in the boring bar assembly, the fixed end of the boring bar in the boring bar assembly is connected to a second mounting end of the guide rail via the floating support, and the working end of the slide rail in the boring bar assembly is connected to the fixed end of the cutting tool. The boring bar assembly includes a cutting tool, a... The device comprises a slide rail, a push rod, a boring bar, a magnetostrictive element, a permanent magnet, a coil, and a mounting plate. The mounting plate is connected to the internal mounting end of the boring bar. The fixed end of the mounting plate is connected to the fixed end of the permanent magnet. The permanent magnet has a coil on it and is distributed in a ring along the axis of the magnetostrictive element. The fixed end of the magnetostrictive element is connected to the first mounting end of the push rod. The second mounting end of the push rod is connected to the fixed end of the slide rail. The working end of the slide rail is connected to the fixed end of the cutting tool.

[0042] Furthermore, it also includes a multi-source data sensor network system, which includes a current sensor, a vibration sensor, an acceleration sensor, and an acoustic emission sensor. The current sensor is mounted on the spindle box, the vibration sensor is mounted at the connection position between the boring bar and the cutting tool, the acoustic emission sensor is mounted in the working area of ​​the cutting tool, and the acoustic emission sensors are evenly spaced along the length of the gun barrel. The acceleration sensors are respectively mounted at the drive end, middle and fixed end of the boring bar, and near the machining hole of the gun barrel.

[0043] Compared with the prior art, the present invention has the following advantages:

[0044] 1. The boring bar and cutting tool of the present invention are equipped with a multi-source data sensing network system to monitor the wear state of the cutting tool, the cutting force, and the dynamic characteristics of the boring bar in real time. Then, the control system adjusts the current to extend or retract the magnetostrictive actuator to change the feed and cutting depth of the cutting tool. Through this micro-displacement high-frequency and high-precision control, the trajectory deviation of the cutting tip is compensated to make the movement trajectory of the cutting tip conform to the expectation. This avoids the serious impact of cutting tip wear and time-varying excitation on the dynamic characteristics of the large length-to-diameter ratio boring bar, which would cause the cutting tip trajectory to change over time. This allows the movement trajectory of the cutting tip to be well controlled, realizing the precision control of the gun barrel, thereby ensuring machining accuracy.

[0045] 2. The cutting tool in this invention is made of high wear-resistant material and is designed with a special cutting edge shape to improve cutting efficiency and extend service life; the boring bar has a large length-to-diameter ratio, which can provide good rigidity and ensure the stability of the tool in deep hole machining.

[0046] 3. In the boring process, the multi-source data sensor network system monitors the wear state and cutting force of the tool in real time. The controller dynamically adjusts the feed rate of the boring bar and the position of the tool based on this information to ensure machining accuracy. By adopting a combination of feedforward predictive control and feedback closed-loop control, machining errors are effectively reduced. Attached Figure Description

[0047] Figure 1 This is a flowchart of the deep hole machining method for artillery barrels based on magnetostrictive actuators according to the present invention;

[0048] Figure 2 This is an overall structural diagram of the deep hole machining device for artillery barrels based on a magnetostrictive actuator according to the present invention;

[0049] Figure 3 This is a schematic diagram of the magnetostrictive actuator in the deep hole machining device for artillery barrels based on the magnetostrictive actuator of the present invention;

[0050] Figure 4 This is a diagram showing the ideal tool tip motion trajectory in the deep hole machining method for artillery barrels based on magnetostrictive actuators of this invention.

[0051] Figure 5 This is a deflection diagram of the boring bar in the deep hole machining method for artillery barrels based on a magnetostrictive actuator according to the present invention.

[0052] Figure 6 This diagram illustrates the barrel machining error caused by boring bar deflection in the deep hole machining method for artillery barrels based on a magnetostrictive actuator, as described in this invention.

[0053] Figure 7 This is a tool tip wear diagram of the deep hole machining method for artillery barrels based on magnetostrictive actuators according to the present invention;

[0054] Figure 8 This diagram illustrates the machining error of the artillery barrel caused by tool tip wear in the deep hole machining method for artillery barrels based on magnetostrictive actuators, as described in this invention.

[0055] Key reference numerals:

[0056] 1. Spindle box, 2. Guide rail, 3. Boring bar assembly, 301. Tool, 302. Slide rail, 303. Push rod, 304. Boring bar, 305. Magnetostrictive element, 306. Permanent magnet, 307. Coil, 308. Mounting plate, 4. Floating support, 5. Cannon barrel, 6. Upper deviation of tool tip movement trajectory, a. Lower deviation of tool tip movement trajectory, b. Shape of inner surface of barrel after actual machining, c. Undeflected boring bar under ideal machining conditions, d. Deflected boring bar under actual machining conditions, e. Shape of barrel after actual machining, f. Shape of barrel after ideal machining, g. Ideal tool condition, h. Worn tool condition, i. Detailed Implementation

[0057] To provide a detailed description of the technical content, objectives, and effects of this invention, the following description will be provided in conjunction with the accompanying drawings.

[0058] Deep hole machining method for artillery barrels based on magnetostrictive actuators, such as Figure 1 As shown, the specific implementation steps are as follows:

[0059] S1. The gun barrel 5 to be processed is mounted on the guide rail 2 through the floating support 4. The cutting tool 301 is mounted on the working end of the boring bar 304. The driving end of the boring bar 304 is connected to the spindle box 1. The fixed end of the boring bar 304 is mounted on the guide rail 2 through the floating support 4.

[0060] S2. Install the multi-source data sensor network system at the corresponding positions of the boring bar 304, the cutting tool 4, and the cannon barrel 5, and set the cutting parameters of the cutting tool 301; the cutting parameters include cutting speed, feed rate, and current intensity.

[0061] Specifically, the multi-source data sensor network system can monitor tool wear and cutting force changes in real time, providing important data support for the controller. The controller receives data from the multi-source data sensor network system and combines it with the tool wear prediction model to dynamically adjust the feed rate of the boring bar 304 and the micro-displacement of the tool 301. Specifically, it includes current sensors, vibration sensors, acceleration sensors, and acoustic emission sensors. The current sensors are mounted on the spindle box 1, the vibration sensors are mounted at the connection point between the boring bar 304 and the tool 301, the acoustic emission sensors are mounted in the working area of ​​the tool 301, and the acoustic emission sensors are evenly spaced along the length of the gun barrel 5. The acceleration sensors are mounted at the drive end, middle, and fixed end of the boring bar 304, and at a position near the machining hole on the gun barrel 5.

[0062] The positional distribution of the accelerometer on the boring bar 304 was obtained by analyzing the sensitivity of the signals from different mounting positions of the accelerometer on the boring bar 304 to the wear of the tool tip 301 using the Pearson correlation coefficient method. The expression for the Pearson correlation coefficient is as follows:

[0063]

[0064] Where, ρ X,Y These are the signals from the accelerometer at different mounting positions on the boring bar 304. X and Y are the correlation coefficients in the Pearson correlation coefficient method, cov(X,Y) is the covariance of X and Y, and σ X and σ Y It is the standard deviation of X and Y.

[0065] S3. Start the spindle box 1 and the multi-source data sensor network system respectively. Process and analyze the multi-source signals collected in real time by the multi-source data sensor network system to construct a sample set. Extract the time-frequency information from the multi-source signals in the sample set. The obtained time-frequency information is used to construct a tool wear prediction model based on neural network deep learning. Specifically, the multi-source signals consist of acoustic emission signals and vibration signals, including tool position deviation values, abnormal cutting force values, abnormal vibration values, and acoustic signal characteristic values.

[0066] S4. Ideally, the tip of tool 301 should follow the set trajectory as shown in the image. Figure 4 The smooth movement shown ensures that the actual machined inner surface shape c of the tube lies between the deviation a on the tool tip's movement trajectory and the deviation b below the tool tip's movement trajectory. Figure 5 As shown, the deflection of the boring bar 304 is formed by the difference between the undeflected boring bar d under ideal machining conditions and the deflected boring bar e under actual machining conditions, as... Figure 7 As shown, the wear of tool 301 is caused by the difference between the ideal tool state h and the worn tool state i. Figure 7 Wear of the 301 medium cutting tool and Figure 5 The deflection of the 304 boring bar will cause the tool tip trajectory to deflect during actual machining. Figure 6 and Figure 8 The offset shown causes a deviation between the actual machined barrel shape f and the ideal machined barrel shape g. Based on the unidirectional long stroke characteristics of the deep hole boring system on the artillery barrel 5, a tool wear prediction model is established.

[0067] S5. Input the acoustic emission signal and vibration signal obtained in step S3 into the tool wear prediction model constructed in step S4, calculate the wear state of tool 301, and generate the current signal of the control system.

[0068] S6. The current signal obtained in step S5 is input to the coil 307 wound around the permanent magnet 306. The magnitude of the current directly determines the magnetic field strength, controlling the extension and retraction of the magnetostrictive element 305, which drives the push rod 303 to perform micron-level displacement, keeping the tool 301 in contact with the cannon barrel 5, thereby achieving fine adjustment of the tool 301's displacement. Acoustic emission signals and vibration signals are continuously collected and monitored to form a closed-loop feedback control system.

[0069] Preferably, the specific implementation steps for step S3, which extracts time-frequency information from multi-source signals, are as follows:

[0070] S31. Analyze the spatiotemporal characteristics and singularities of multi-source signals. Based on the characteristics of tool wear and machining conditions, use a sliding frame sampling method to extract signal data of a specified length to ensure that the extracted signal data is sufficient to capture the key temporal characteristics of tool wear each time. Analyze the dispersion of the signal data, remove outlier data points that exceed or fall below three standard deviations of the average value, and obtain the initial data sample.

[0071] Specifically, for each data point, its standardized value z is calculated to determine whether it is outlier. The formula is as follows:

[0072]

[0073] Where z is the standardized value, X is the signal data, μ is the mean of the signal dataset, and σ is the standard deviation of the signal dataset. If |z|>3, the data is considered an outlier. This method can remove outliers that are more than or less than three times the standard deviation of the mean, ensuring high accuracy of the signal data.

[0074] S32. Based on the initial data samples, construct a single-condition source domain data sample set using the tool tip wear value as a label.

[0075] S33. Based on the single-condition source domain data sample set obtained in step S32, its time domain information is acquired. The low-frequency and high-frequency components of the single-condition source domain data sample set are simultaneously decomposed using wavelet transform to obtain frequency domain information. The expression for the wavelet transform method is:

[0076]

[0077] Among them, W a,b (t) represents the time-frequency local characteristics of the processing vibration, acceleration, and acoustic emission signals obtained after wavelet transform at different scales and locations, and x(t) represents the original signals acquired by multiple sensors during the actual processing. The complex conjugate of the mother wavelet function used in the wavelet transform is represented by , and a and b represent different levels and time characteristics of tool wear, respectively.

[0078] Wavelet transform can accurately extract the time and frequency domain information of a signal, and in the processing method of this invention, it can identify instantaneous features related to the wear of the tool 301.

[0079] S34. The frequency domain information obtained in S33 is processed using the Hilbert-Huang method to extract the instantaneous frequency components of the vibration frequency change and cutting force fluctuation caused by tool wear, achieving high-resolution time-frequency analysis and obtaining the time-frequency information from the multi-source signal. The expression for the Hilbert-Huang method is:

[0080]

[0081] Wherein, ω(t) represents the instantaneous frequency characteristics of the signal caused by the change in the contact characteristics between the tool 301 and the cannon barrel 5, and θ(t) represents the phase characteristics of the vibration of the tool 301 and the machining force as the wear degree changes.

[0082] Preferably, the process of establishing the tool wear prediction model in step S4 is as follows:

[0083] S41. First, deep autoencoder technology is used to extract multi-scale spatiotemporal features of the time-frequency information obtained in step S3. Correlation analysis and kernel principal component analysis (KPCA) are used to reduce the dimensionality of the multi-scale spatiotemporal features. Then, a deep bidirectional recurrent neural network (Bi-LSTM) method is used, with an attention mechanism introduced to adaptively calculate and adjust the input weights. The nonlinear mapping relationship between the features of multi-source signals and the tool tip wear sensitivity features of tool 301 is explored. A tool tip wear prediction model under a single working condition is established using the time-series signal features.

[0084] The expressions for correlation analysis and KPCA methods are as follows:

[0085]

[0086] Where φ(X) is the feature representation mapped to a high-dimensional space through the kernel function K, and α i These are the projection coefficients. The state update expression for a bidirectional recurrent neural network is:

[0087] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0088] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0089]

[0090] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0091] h t =o t *tanh(C t )

[0092] Among them, ft This indicates the forgetting of historical signal data during processing, i t C represents the acceptance of new signal data during processing. t This indicates the long-term storage and memory of signal data during the processing, which is updated at each time step. This indicates that the data generated based on the current input data is used to update C. t The information of the state, h t This indicates adjustments to the current forecast during the processing. t For the final data output, x t This represents the signal collected from the multi-source data sensor network system at the current moment, σ represents the Sigmoid activation function, ensuring that the signal output value is between (0,1), and W f W represents the weight matrix of the forget gate. i W represents the weight matrix of the input gate. C W represents the weight matrix for calculating the processing state update. o Let b represent the weight matrix of the output gate. f The bias term representing the forget gate, b i The input gate bias term, b C b represents the bias term used to update the current memory cell state. o The bias term h of the output gate t-1 C represents the hidden state at the previous moment. t-1 This indicates the stored state of the tool displacement compensation at the previous moment, and tanh represents the hyperbolic tangent activation function, which normalizes the signal data to the range (-1, 1).

[0093] S42. To improve the generalization ability of the model, based on the tool tip wear prediction model under a single working condition obtained in step S41, a feature extractor is constructed using a single working condition source domain data sample set as input and a target domain to be learned under the new working condition as output. This feature extractor uses a convolutional neural network (CNN) and a bidirectional long short-term memory network (BLSTM) to extract multi-dimensional spatiotemporal features of the single working condition source domain data sample set and the target domain to be learned under the new working condition, thus constructing a domain-adaptive transfer model. The domain-adaptive transfer model mainly includes a feature extractor composed of CNN and BLSTM, a tool wear classifier composed of FC layers, and a metric function—Maximum Mean Discrepancy (MMD). Labeled samples from the source domain and unlabeled samples from the target domain are first input into the feature extractor. Through multi-source data fusion and nonlinear mapping feature modeling of sequence features, high-dimensional feature vectors are extracted simultaneously from the source and target domain samples. Multiple-kernel MMD (MK-MMD) is used to quantize and minimize the distribution differences of transferable features between domains to obfuscate the spatial distribution differences of features between the source and target domains, thereby achieving knowledge transfer and sharing and improving the generalization ability of the prediction model.

[0094] In a preferred embodiment of the present invention, a deep hole machining apparatus for artillery barrels based on a magnetostrictive actuator is provided, such as... Figure 2 As shown, the assembly includes a spindle box 1, a guide rail 2, a boring bar assembly 3, and a floating support 4. The fixed end of the spindle box 1 is connected to the first mounting end of the guide rail 2. The spindle hole of the spindle box 1 is connected to the positioning component of the boring bar 304 in the boring bar assembly 3. The fixed end of the boring bar 304 in the boring bar assembly 3 is connected to the second mounting end of the guide rail 2 through the floating support 4. The working end of the slide rail 302 in the boring bar assembly 3 is connected to the fixed end of the tool 301. The floating support 4 can move along the guide rail 2, and the boring bar 304 can move axially along the floating support 4, thereby driving the tool 301 to move axially along the boring bar 304. The tool 301 can extend and retract in a direction perpendicular to the axial direction of the boring bar 304 under the drive of the magnetostrictive element 305, thereby realizing variable diameter boring.

[0095] Boring bar assembly 3, such as Figure 3As shown, the assembly includes a cutting tool 301, a slide rail 302, a push rod 303, a boring bar 304, a magnetostrictive element 305, a permanent magnet 306, a coil 307, and a mounting plate 308. The mounting plate 308 is connected to the mounting end inside the boring bar 304. The fixed end of the mounting plate 308 is connected to the fixed end of the permanent magnet 306. The permanent magnet 306 is provided with a coil 307. The permanent magnet 306 is distributed in a ring along the axis of the magnetostrictive element 305. The fixed end of the magnetostrictive element 305 is connected to the first mounting end of the push rod 303. The second mounting end of the push rod 303 is connected to the fixed end of the slide rail 302. The working end of the slide rail 302 is connected to the fixed end of the cutting tool 301. The slide rail 302 can convert the axial movement of the push rod 303 into the longitudinal movement of the cutting tool 301. High-frequency micro-displacement control is achieved through current adjustment. The magnetostrictive element 305 ensures the precise adjustment of the cutting tool 301 during the machining process.

[0096] The magnetostrictive element 305 is located inside the boring bar 304. It achieves high-frequency micro-displacement control through current regulation to ensure the precise adjustment of the tool 301 during the machining process. The control system is connected to various sensors in the multi-source data sensor network system, including the data acquisition module and the control module, which are responsible for real-time monitoring of the machining status and adjusting the control strategy. The magnetostrictive element 305 is connected to the control system through a cable to transmit current signals to regulate the output of the magnetostrictive element 305.

[0097] A magnetic field is generated around the magnetostrictive element 305 by the coil 307 and the permanent magnet 306, causing the magnetostrictive element 305 to shorten or lengthen under the influence of the magnetic field. The coil 307 is controlled by the current control module, and the control system of the processing device of the present invention adjusts the current intensity in the coil 307 through the current control module, thereby controlling the intensity of the magnetic field. The magnetostrictive element 305 stretches under the action of the magnetic field, thereby adjusting the axial movement of the push rod 303 inside the boring bar 304 in real time, pushing the push rod 303 to a micron-level displacement, so as to precisely control the feed and cutting depth of the tool 301.

[0098] A current sensor is located in the spindle system power path, monitoring spindle current changes in real time, assessing cutting force and tool load status, and providing early warning of tool 301 wear or cutting abnormalities. A vibration sensor is located in the clamping area of ​​the boring bar 304 and tool 301, monitoring the vibration of the boring bar 304 and tool 301 in real time, identifying potential vibration problems and tool 301 wear status, and optimizing machining stability. An acoustic emission sensor is located in the tool 301 clamping area, capturing acoustic emission signals during the cutting process, and analyzing the tool 301 wear status and cutting quality through acoustic signal analysis. In the multi-source data sensor network system, each sensor monitors data in real time and converts it into electrical signals, which are then transmitted wirelessly to the control system's data acquisition module for data acquisition based on the OPC UA protocol. The control system is responsible for real-time monitoring of machining status and adjusting control strategies.

[0099] The following describes in further detail a method and apparatus for deep hole machining of artillery barrels based on magnetostrictive actuators according to the present invention, with reference to specific embodiments:

[0100] First, the cannon barrel 5 is mounted on the guide rail 2 via the floating support 4. The position of the cannon barrel 5 is adjusted to ensure that the cutting trajectory of the tool 301 is aligned with the cannon barrel 5 during machining. The multi-source signals collected by the multi-source data sensor network system are input to the data acquisition module to ensure real-time monitoring of the machining status. The control system is started, and initial parameters are set, including cutting speed, feed rate, and current intensity. The current magnitude of the coil 307 on the permanent magnet 306 is adjusted by the control system to change the magnetic field strength, causing the magnetostrictive element 305 to extend or retract, making the push rod 303 move forward and backward, thereby adjusting the tool 301 to extend and retract along the axis perpendicular to the boring bar 304. During the boring process, the tool tip wear, vibration, and current data of the tool 301 are collected in real time.

[0101] Next, a tool wear prediction model was established. Considering the unidirectional long stroke characteristics of the deep hole boring system for the gun barrel 5, the signal from the current sensor located on the spindle box 1 in the multi-source data sensor network system was read by the boring machine CNC system. Accelerometers were installed on different parts such as the boring bar 304 and the gun barrel 5, and acoustic emission sensors were evenly spaced along the length of the gun barrel 5. Thus, the multi-source data sensor network system was used to acquire multi-source signals. The spatiotemporal characteristics and singularities of the multi-source signals were analyzed, and a sliding frame sampling method was used to extract signal data of a specified length. The dispersion was analyzed, and outliers exceeding or falling below three standard deviations of the average value were removed to obtain the initial data sample.

[0102] Then, using multi-source signals as data samples and the tool tip wear value of tool 301 as the label, a single-condition source domain data sample set is constructed. Based on the above source domain data samples, its time domain information is obtained. Wavelet transform is used to simultaneously decompose the denoised low-frequency and high-frequency parts to obtain frequency domain information. Then, the Hilbert-Huang method is used to obtain the instantaneous frequency components with actual physical meaning in the signal, thereby achieving high-resolution time-frequency analysis. Multi-scale spatiotemporal features in the time-frequency information of multi-source signals are extracted through deep autoencoder technology, and feature dimensionality reduction is achieved using correlation analysis and KPCA. A tool wear prediction model based on a deep bidirectional recurrent neural network under a single-condition is established. An attention mechanism is introduced to adaptively calculate and adjust the input weights, and the nonlinear mapping relationship between multi-source signal feature fusion and tool tip wear sensitive features in tool 301 is explored. Using the source domain samples of the original condition as input and the target domain to be learned in the new condition as output, a domain adaptive transfer model is constructed.

[0103] Finally, real-time analysis is performed using a preset tool wear model. Based on the analysis results, the control strategy is dynamically adjusted and the current output is optimized to achieve high-frequency micro-displacement control of tool 301, ensuring the stability of the tool tip throughout the machining process. After boring, the equipment is shut down, and the machining data is extracted for analysis and improvement. The machining quality is evaluated, and the straightness, roundness, and inner hole dimensions of the deep hole are checked to ensure that design requirements are met.

[0104] like Figure 4 As shown, in the ideal machining process, the tip of the tool 301 moves stably along a predetermined trajectory, and within the allowable tolerance range, the contact surface between the tool 301 and the gun barrel 5 remains stable, ensuring machining accuracy; as Figure 5 As shown, during high-speed cutting, the boring bar 304 will deflect to a certain extent due to the cutting force and changes in machining conditions. This deflection directly affects the movement trajectory of the tool 301, especially in deep hole machining, where the deflection effect is particularly pronounced due to the relatively long length of the boring bar 304. Figure 6 As shown, due to the cumulative effect of the boring bar deflection, the movement trajectory of the tool 301 deviates from the ideal trajectory, resulting in an irregular machined surface, which in turn reduces the machining accuracy of the deep hole of the cannon barrel 5, with a machining error of Δ1; Figure 7 As shown, during the contact process between the cutting tool 301 and the cannon barrel 5, due to material cutting and heat accumulation, the tip of the cutting tool 301 will experience a certain degree of wear. This wear gradually intensifies with the continuation of the machining process. As the wear increases, the shape and position of the cutting tool tip change, directly leading to deviations in the cutting tool 301's trajectory. Figure 8As shown, due to the wear of the tool tip 301, the movement trajectory of the tool 301 deviates significantly, resulting in errors in deep hole machining. This error is mainly manifested in the change of the tool tip position, causing the tool 301 to be unable to cut along the ideal trajectory during machining, thus resulting in a dimensional deviation Δ1 in the deep hole machining of the cannon barrel 5. To compensate for the errors caused by the deflection of the boring bar 304 and the wear of the tool tip 301, a multi-source data sensor network system is used to monitor the dynamic signals during machining in real time, especially the state of the tool 301 and the vibration signal of the boring bar 304. By acquiring acoustic emission and vibration signals in real time, the control system adjusts the current of the coil 307 on the permanent magnet 306 in real time according to the returned signals, thereby changing the magnetic field strength, causing the magnetostrictive element 305 to extend and retract, which in turn drives the axial movement of the push rod 303, causing the tool 301 to adjust the extension and retraction displacement L along the slide rail 302, thereby compensating for the trajectory deviation of the tool tip 301 and making the tool tip movement trajectory of the tool 301 conform to the ideal expectation, thus ensuring machining accuracy.

[0105] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for deep hole machining of artillery barrels based on magnetostrictive actuators, characterized in that, It includes the following steps: S1. The large length-to-diameter ratio gun barrel to be machined is mounted on the guide rail by a floating support. The cutting tool is mounted on the working end of the boring bar. The driving end of the boring bar is connected to the spindle box. The fixed end of the boring bar is mounted on the guide rail by a floating support. S2. Install the multi-source data sensor network system at the corresponding positions on the boring bar, cutting tool, and gun barrel, and set the cutting parameters of the cutting tool. S3. Start the spindle box and the multi-source data sensor network system respectively, process the acoustic emission signal and vibration signal collected in real time by the multi-source data sensor network system, and extract the time-frequency information about the tool tip displacement from the acoustic emission signal and vibration signal. S4. Based on the characteristics of deep hole machining of artillery barrels, establish a tool wear prediction model, which includes the following sub-steps: S41. Using the state update expression of a bidirectional recurrent neural network, with the acoustic emission signal and vibration signal collected in step S3 as the core input and the tool tip displacement as the output, a tool tip wear prediction model under a single working condition is established. The nonlinear mapping relationship between the characteristics of multi-source signals and the tool tip wear-sensitive characteristics is explored. The state update expression of the bidirectional recurrent neural network is as follows: f t =σ(W f ·[h t-1 ,x t ]+b f ) i t =σ(W i ·[h t-1 ,x t ]+b i ) the t =σ(W o ·[h t-1 ,x t ]+b o ) h t =o t *fishy(C) t ) Among them, f t This indicates the forgetting of historical signal data during processing, i t C represents the acceptance of new signal data during processing. t This indicates the long-term storage and memory of signal data during the processing, which is updated at each time step. This indicates that the data generated based on the current input data is used to update C. t The information of the state, h t This indicates adjustments to the current forecast during the processing. t For the final data output, x t This represents the signal collected from the multi-source data sensor network system at the current moment, σ represents the Sigmoid activation function, ensuring that the signal output value is between (0,1), and W f W represents the weight matrix of the forget gate. i W represents the weight matrix of the input gate. C W represents the weight matrix for calculating the processing state update. o Let b represent the weight matrix of the output gate. f The bias term representing the forget gate, b i The input gate bias term, b C b represents the bias term used to update the current memory cell state. o The bias term h of the output gate t-1 C represents the hidden state at the previous moment. t-1 This indicates the stored state of the tool displacement compensation at the previous moment, and tanh represents the hyperbolic tangent activation function, which normalizes the signal data to the range (-1, 1). S42. Based on the tool tip wear prediction model under a single working condition obtained in step S41, the single working condition source domain data sample set is used as input, and the target domain to be learned under the variable feed working condition is used as output. A feature extractor is constructed using a convolutional neural network and a bidirectional long short-term memory network to extract the multi-dimensional spatiotemporal features of the single working condition source domain data sample set and the target domain to be learned under the variable feed working condition, and a domain adaptive transfer model is constructed. S5. Input the time-frequency information of the acoustic emission signal and vibration signal obtained in step S3 into the tool wear prediction model constructed in step S4, calculate the tool wear state, and generate the current signal of the control system. S6. Input the current signal obtained in step S5 into the coil wound on the permanent magnet to change the magnetic field strength, thereby controlling the extension and retraction of the magnetostrictive element, driving the push rod to make a micron-level displacement, so that the tool tip and the cannon barrel are kept in contact, thereby realizing the fine adjustment of the tool displacement, continuously collecting and monitoring acoustic emission signals and vibration signals to form a closed-loop feedback control.

2. The method for deep hole machining of artillery barrels based on magnetostrictive actuators according to claim 1, characterized in that, The distribution of accelerometer positions on the boring bar was obtained by analyzing the sensitivity of signals from different mounting positions of the accelerometers on the boring bar to the tool tip wear using the Pearson correlation coefficient method. The expression for the Pearson correlation coefficient is as follows: Where, ρ X,Y These are the signals from the accelerometer at different mounting positions on the boring bar. X and Y are the correlation coefficients in the Pearson correlation coefficient method, cov(X,Y) are the covariances of X and Y, and σ is the signal from the accelerometer at different mounting positions on the boring bar. X and σ Y These are the standard deviations of X and Y, respectively.

3. The method for deep hole machining of artillery barrels based on magnetostrictive actuators according to claim 1, characterized in that, In step S2, the cutting parameters include cutting speed, feed rate, and current intensity; in step S3, the multi-source signals include tool position deviation value, abnormal cutting force value, abnormal vibration value, and acoustic signal characteristic value.

4. The method for deep hole machining of artillery barrels based on magnetostrictive actuators according to claim 3, characterized in that, The specific implementation steps for step S3, extracting time-frequency information from multi-source signals, are as follows: S31. Based on the characteristics of tool wear and machining conditions, a sliding frame is used to sample and extract signal data of a specified length. The dispersion of the signal data is analyzed, and outlier data points that exceed or fall below three times the standard deviation of the average value are removed to obtain the initial data sample. S32. Based on the initial data samples, construct a single-condition source domain data sample set using the tool tip wear value as a label; S33. Based on the single-condition source domain data sample set obtained in step S32, the low-frequency and high-frequency components of the single-condition source domain data sample set are simultaneously decomposed using wavelet transform to obtain frequency domain information. The expression for the wavelet transform method is: Among them, W a,b (t) represents the time-frequency local features of the processing vibration, acceleration, and acoustic emission signals at different scales and locations after wavelet transform, and x(t) represents the original signals acquired by multiple sensors during the actual processing. The complex conjugate of the mother wavelet function used in the wavelet transform is represented by , and a and b represent different levels and time characteristics of tool wear, respectively. S34. The frequency domain information obtained in S33 is processed using the Hilbert-Huang method to extract the instantaneous frequency components of the vibration frequency change and cutting force fluctuation caused by tool wear, thus obtaining the time-frequency information in the multi-source signal. The expression for the Hilbert-Huang method is as follows: Where ω(t) represents the instantaneous frequency characteristics of the signal caused by the change in the contact characteristics between the tool and the gun barrel, and θ(t) represents the phase characteristics of the tool vibration and machining force as the wear degree changes.

5. The method for deep hole machining of artillery barrels based on magnetostrictive actuators according to claim 4, characterized in that, In step S31, the judgment expression for the abnormal data point is: Where z is the standardized value, X is the signal data, μ is the mean of the signal dataset, σ is the standard deviation of the signal dataset, and if |z|>3, the data is considered an outlier.

6. A deep hole machining device for artillery barrels based on a magnetostrictive actuator, characterized in that, It includes a spindle box, guide rails, boring bar assembly, and floating support. The fixed end of the spindle box is connected to the first mounting end of the guide rail, the output end of the spindle box is connected to the drive end of the boring bar in the boring bar assembly, the fixed end of the boring bar in the boring bar assembly is connected to the second mounting end of the guide rail through a floating support, and the working end of the slide rail in the boring bar assembly is connected to the fixed end of the tool; the boring bar assembly includes a tool, a slide rail, a push rod, a boring bar, a magnetostrictive element, a permanent magnet, a coil, and a mounting plate, the mounting plate is connected to the mounting end inside the boring bar, the fixed end of the mounting plate is connected to the fixed end of the permanent magnet, the permanent magnet is provided with a coil, the permanent magnet is distributed in a ring along the axis of the magnetostrictive element, the fixed end of the magnetostrictive element is connected to the first mounting end of the push rod, the second mounting end of the push rod is connected to the fixed end of the slide rail, and the working end of the slide rail is connected to the fixed end of the tool.

7. The deep hole machining device for artillery barrels based on a magnetostrictive actuator according to claim 6, characterized in that, It also includes a multi-source data sensor network system, which includes a current sensor, a vibration sensor, an acceleration sensor and an acoustic emission sensor. The current sensor is mounted on the spindle box, the vibration sensor is mounted at the connection position between the boring bar and the tool, the acoustic emission sensor is mounted in the working area of ​​the tool, and the acoustic emission sensors are evenly spaced along the length of the gun barrel. The acceleration sensors are respectively mounted at the drive end, middle and fixed end of the boring bar and at the position of the gun barrel near the machining hole.

Images