Machine tool vibration detection method and system

Abstract

The invention relates to the technical field of machine tool vibration detection, and discloses a machine tool vibration detection method and system, and the method comprises the steps: collecting vibration signals of all regions, and forming a first multi-source independent signal data set; carrying out de-aliasing processing on the partially aliasing vibration signals; extracting time-frequency characteristics of machine tool vibration; constructing a feature vector set according to the extracted features, and inputting the feature vector set into a machine tool vibration classification model; and quantifying the contribution proportion of each vibration source. Compared with the prior art that only vibration signal acquisition and simple analysis can be carried out on a single area of a machine tool, and particularly under the conditions of complex vibration signal sources and multi-source aliasing, accurate separation and recognition of vibration sources cannot be realized, the method and the device have the advantages that the signal de-aliasing and the dynamic convolutional neural network classification model are combined, so that the vibration sources can be accurately separated and recognized; accurate acquisition, separation, analysis and identification of multi-area vibration signals of the machine tool are realized, so that the problem of inaccurate detection results is avoided, and the vibration detection accuracy of the machine tool is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of machine tool vibration detection, and in particular relates to a machine tool vibration detection method and system. Background Art

[0002] At present, during the operation of machine tools, vibration problems are an important factor affecting machining accuracy, surface quality and equipment reliability. The vibration of machine tools may come from multiple vibration sources such as spindle imbalance, abnormal tool cutting, unstable table clamping and base environment resonance. These vibration sources are often superimposed in complex operating environments, making it difficult to distinguish the source of the vibration signal; in addition, vibration problems will cause the surface roughness of the machined workpiece to increase, the dimensional accuracy to decrease, and even accelerate the wear of machine tool components. In severe cases, it may cause machine tool operation failures and shorten the service life of the equipment; in the prior art, the detection method for machine tool vibration signals has the following deficiencies: on the one hand, traditional vibration detection methods usually use a single sensor to monitor a certain area of ​​the machine tool, and fail to achieve comprehensive coverage of vibration signals in multiple areas of the machine tool, and therefore cannot fully reflect the overall vibration characteristics of the machine tool; on the other hand On the one hand, the sources of vibration signals are complex, and multi-source signals are prone to aliasing during collection. The existing technology has limited capabilities in signal separation and de-aliasing, and it is difficult to extract effective vibration features from aliased signals. In addition, vibration signals are often accompanied by high-frequency noise, and the existing signal denoising methods may lose some key signal features while reducing noise interference, thereby affecting the accuracy and reliability of vibration detection. Therefore, there is an urgent need for a method that can collect multi-source vibration signals while covering the machine tool spindle, tool, worktable and base area, and combine signal de-aliasing and denoising technology to achieve accurate separation and identification of multi-source signals in complex environments, thereby improving the comprehensiveness and accuracy of vibration detection, and providing a reliable basis for machine tool fault diagnosis and operation optimization. Summary of the invention

[0003] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to propose a method for detecting machine tool vibration, aiming to solve the technical problem that the prior art can only collect and simply analyze vibration signals in a single area of ​​the machine tool, especially under the conditions of complex vibration signal sources and multi-source aliasing, and cannot achieve accurate separation and identification of vibration sources.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides a method for detecting vibration of a machine tool.

[0005] The method for detecting machine tool vibration comprises:

[0006] Step S10: installing multi-source sensors in the spindle area, tool area, worktable area and base area of ​​the target machine tool, including: a spindle area sensor, installed on the spindle bearing housing, used to record the unbalanced vibration of high-speed rotation; a tool area sensor, installed on the tool holder, used to collect cutting vibration and tool abnormality signals; a worktable area sensor, installed on the worktable base, used to capture vibration caused by worktable movement or unstable clamping of workpieces; a base area sensor, installed on the machine tool base, used to detect resonance interference transmitted to the machine tool by external vibration;

[0007] The vibration signals of various areas of the target machine tool are collected in real time by multi-source sensors to form a first multi-source independent signal data set S collected ={S spindle ,S tool ,S workbench ,S base}, where S spindle is the vibration signal of the main axis area, S tool is the vibration signal of the tool area, S workbench is the vibration signal of the workbench area, S base is the vibration signal of the base area;

[0008] Step S20: Vibration signal S for the tool area tool and the vibration signal S of the workbench area workbench Performing a de-aliasing process on the second multi-source independent signal data set to form a second multi-source independent signal data set, and performing a denoising process on the second multi-source independent signal data set to generate a third multi-source independent signal data set S optimized , in, is the denoised signal in the spindle region, is the signal after de-aliasing in the tool area, is the signal after de-aliasing in the workbench area, is the denoised signal of the base area;

[0009] Step S30: by performing a third multi-source independent signal data set S optimized Perform short-time Fourier transform to extract the time-frequency feature distribution The formula is:

[0010]

[0011] in, is the third multi-source independent signal dataset S optimized Any signal h(τ-t) is the Hamming window function with 128 sampling points; is the time-frequency distribution of the signal; t is the time center point of the current signal, f is the center value of the frequency component, and τ is the integral variable of time, which is used to traverse the entire time range of the signal;

[0012] From the time-frequency feature distribution Extract the key characteristics of machine tool vibration, including the main frequency component f main , vibration energy distribution E(f) and vibration amplitude dynamic range A range ;

[0013] is the third multi-source independent signal dataset S optimized middle Any signal constructs a feature vector, each of which contains the corresponding vibration main frequency component, vibration energy distribution and vibration amplitude dynamic range, and generates a feature vector set V optimized = {V spindle ,V tool ,V workbench ,V base}, where V spindle ,V tool ,V workbench ,V base They are the feature vector of the spindle area, the feature vector of the tool area, the feature vector of the worktable area and the feature vector of the base area;

[0014] Step S40: Construct a machine tool vibration classification model, the classification model input is a feature vector set V optimized ,The classification model outputs the categories of vibration factors, including spindle unbalance vibration, tool cutting vibration, worktable clamping vibration and base environment interference vibration;

[0015] Step S50: quantify the contribution ratio of the vibration source corresponding to the output vibration factor category and generate a detection report.

[0016] Preferably, in step S10, the multi-source sensor uses a three-axis acceleration sensor, model "ADXL356", with a measuring range of ±40g, a sensitivity of 10mV / g, and a sampling frequency set to 20kHz.

[0017] Preferably, in step S20, the vibration signal S of the tool area tool and the vibration signal S of the workbench area workbench To perform anti-aliasing, use the formula:

[0018]

[0019] is the signal after de-aliasing of the tool area sensor, k spindle is the interference coefficient of the vibration signal of the spindle vibration on the tool area, and the value range is [0, 1];

[0020]

[0021] is the signal after de-aliasing of the sensor in the workbench area, k base is the interference coefficient of the base vibration on the vibration signal of the workbench area.

[0022] Preferably, in step S30, the vibration main frequency component f maim The calculation formula is in, for The value of f at maximum; the calculation formula of vibration energy distribution E(f) is Vibration amplitude dynamic range A range The calculation formula is A range =A max -A min , where A max and A min are respectively the maximum amplitude and minimum amplitude of the third multi-source independent signal.

[0023] Preferably, in step S40, the step of outputting the vibration factor category by the classification model includes:

[0024] The eigenvalue F is extracted through the dynamic convolution layer conv (i,j), the calculation formula is:

[0025]

[0026] Among them, F conv (i, j) represents the feature value at position (i, j) in the output feature map after the convolution operation, W ′ is the dynamically adjusted convolution kernel weight, k is the size of the convolution kernel, C is the number of input channels, and b conv is the bias term, m, n represent the position offset of the convolution kernel in the vertical and horizontal directions respectively, and c is the input feature map channel;

[0027] The probability distribution of the final classification is obtained through the fully connected layer and the Softmax function. The calculation formula is:

[0028] P = Softmax(W fc ·F conv (i,j)+b fc )

[0029] Among them, P is the probability distribution of the vibration signal belonging to each category, W fc is the weight matrix of the fully connected layer, b fc is the bias parameter of the fully connected layer;

[0030] Output the vibration factor category according to the value of P.

[0031] Preferably, the dynamic parameter W ′ The adjustment of is dynamically adjusted by inputting the feature vector, and its formula is:

[0032] W ′ =W·σ(G)

[0033] G = ReLU(W g ·V+b g ),

[0034] Where W is the initial convolution kernel weight matrix; G is the dynamic adjustment factor generated by the input feature vector, which is calculated using the following formula: W ′ =W·σ(ReLU(W g ·V+b g )), where V is the input feature vector, including the main frequency component, amplitude dynamic range and spectrum energy; W g is a preset linear transformation matrix used to map the input feature vector to the space of dynamic adjustment factors; b g is the bias vector; ReLU is the activation function used to ensure that the output is non-negative; σ is the Sigmoid activation function, which is used to normalize the adjustment factor to the range of [0, 1].

[0035] Preferably, in step S50, the vibration source contribution ratio is calculated by the ratio of the spectrum energy of the vibration source.

[0036] The present invention also provides a machine tool vibration detection system comprising:

[0037] The vibration signal acquisition module is used to install multi-source sensors in the spindle area, tool area, worktable area and base area of ​​the target machine tool, including: a spindle area sensor installed on the spindle bearing housing to record the unbalanced vibration of high-speed rotation; a tool area sensor installed on the tool holder to collect cutting vibration and tool abnormality signals; a worktable area sensor installed on the worktable base to capture vibration caused by worktable movement or unstable clamping of workpieces; a base area sensor installed on the machine tool base to detect resonance interference transmitted to the machine tool by external vibration;

[0038] The vibration signals of various areas of the target machine tool are collected in real time by multi-source sensors to form a first multi-source independent signal data set S collected ={S spindle ,S tool ,S workbench ,S base}, where S spindle is the vibration signal of the main axis area, S toolis the vibration signal of the tool area, S workbench is the vibration signal of the workbench area, S base is the vibration signal of the base area;

[0039] Signal de-aliasing and denoising module for vibration signal S in the tool area toil and the vibration signal S of the workbench area workbench Performing a de-aliasing process on the second multi-source independent signal data set to form a second multi-source independent signal data set, and performing a denoising process on the second multi-source independent signal data set to generate a third multi-source independent signal data set S optimized , in, is the denoised signal in the spindle region, is the signal after de-aliasing in the tool area, is the signal after de-aliasing in the workbench area, is the denoised signal of the base area;

[0040] The feature extraction module is used to extract the third multi-source independent signal data set S optimized Perform short-time Fourier transform to extract the time-frequency feature distribution The formula is:

[0041]

[0042] in, is the third multi-source independent signal dataset S optimized Any signal h(τ-t) is the Hamming window function with 128 sampling points; is the time-frequency distribution of the signal; t is the time center point of the current signal, f is the center value of the frequency component, and τ is the integral variable of time, which is used to traverse the entire time range of the signal;

[0043] From the time-frequency feature distribution Extract the key characteristics of machine tool vibration, including the main frequency component f main , vibration energy distribution E(f) and vibration amplitude dynamic range A range ;

[0044] is the third multi-source independent signal dataset S optimized middle Any signal constructs a feature vector, each of which contains the corresponding vibration main frequency component, vibration energy distribution and vibration amplitude dynamic range, and generates a feature vector set V optimized = {V spindle ,V tool ,V worlbemch ,V nase}, where V spindle ,V tool,V workbench ,V base They are the feature vector of the spindle area, the feature vector of the tool area, the feature vector of the worktable area and the feature vector of the base area;

[0045] The vibration factor classification module is used to build a machine tool vibration classification model. The input of the classification model is the feature vector set V optimized ,The classification model outputs the categories of vibration factors, including spindle unbalance vibration, tool cutting vibration, worktable clamping vibration and base environment interference vibration;

[0046] The test report output module is used to quantify the contribution ratio of the vibration source corresponding to the output vibration factor category and generate a test report.

[0047] The present invention also provides a machine tool vibration detection device, comprising a memory, a processor and a machine tool vibration detection program stored in the memory and executable on the processor, wherein the machine tool vibration detection program implements the machine tool vibration detection method when executed by the processor.

[0048] The present invention also provides a computer program product, including a machine tool vibration detection program, wherein the machine tool vibration detection method is implemented when the machine tool vibration detection program is executed by a processor.

[0049] The beneficial effect of the present invention is that compared with the prior art which can only collect and simply analyze vibration signals of a single area of ​​a machine tool, especially when the vibration signal sources are complex and multiple sources are mixed, the technical problem of being unable to accurately separate and identify the vibration sources can be solved. This application combines signal de-aliasing with a dynamic convolutional neural network classification model to achieve accurate collection, separation, analysis and identification of vibration signals in multiple areas of a machine tool, thereby avoiding the problem of inaccurate detection results and improving the accuracy of machine tool vibration detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 The figure is a flow chart of a first embodiment of a method for detecting vibration of a machine tool according to the present invention.

[0052] Figure 2 The present invention is a schematic diagram of a device for detecting a machine tool vibration. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Embodiment 1: Figure 1 , which is a flow chart of a first embodiment of a method for detecting machine tool vibration of the present invention, provides a first embodiment of a method for detecting machine tool vibration of the present invention.

[0055] In a first embodiment, the method for detecting vibration of a machine tool includes:

[0056] Step S10: installing multi-source sensors in the spindle area, tool area, worktable area and base area of ​​the target machine tool, including: a spindle area sensor, installed on the spindle bearing housing, used to record the unbalanced vibration of high-speed rotation; a tool area sensor, installed on the tool holder, used to collect cutting vibration and tool abnormality signals; a worktable area sensor, installed on the worktable base, used to capture vibration caused by worktable movement or unstable clamping of workpieces; a base area sensor, installed on the machine tool base, used to detect resonance interference transmitted to the machine tool by external vibration;

[0057] The vibration signals of various areas of the target machine tool are collected in real time by multi-source sensors to form a first multi-source independent signal data set S collected ={S spindle ,S tool ,S workbench ,S hase}, where S spindle is the vibration signal of the main axis area, S tool is the vibration signal of the tool area, S workbench is the vibration signal of the workbench area, S base is the vibration signal of the base area;

[0058] It should be noted that, in the first multi-source independent signal data set, the vibration characteristic corresponding to each signal is a time domain acceleration signal collected by sensors in each region, and the specific form of the signal is time series data.

[0059] Understandably, S spindle It is the vibration signal collected by the sensor in the spindle area, which is mainly composed of the unbalanced vibration component caused by the spindle rotation; S tool It is the vibration signal collected by the sensor in the tool area, including cutting vibration signal and tool abnormality signal; S workbenchIt is the vibration signal collected by the sensor in the workbench area, which mainly reflects the unstable clamping vibration during the movement of the workbench; S base It is the vibration signal collected by the sensor in the base area, which is mainly composed of the resonance caused by the environmental interference transmitted to the base; these signals remain independent in the data set, and the sampling frequency and number of sampling points of each signal are consistent to ensure the timing consistency and calculation accuracy of subsequent de-aliasing and denoising processing. The construction of the first multi-source independent signal data set is the basis for analyzing and classifying the vibration characteristics of different regions.

[0060] Step S20: Vibration signal S for the tool area tool and the vibration signal S of the workbench area workbench Performing a de-aliasing process on the second multi-source independent signal data set to form a second multi-source independent signal data set, and performing a denoising process on the second multi-source independent signal data set to generate a third multi-source independent signal data set S optimized , in, is the denoised signal in the spindle region, is the signal after tool area de-aliasing, is the signal after de-aliasing in the workbench area, is the denoised signal of the base area;

[0061] It should be noted that the vibration signal S for the tool area tool and the vibration signal S of the workbench area workbench To perform anti-aliasing, use the formula:

[0062]

[0063] is the signal after de-aliasing of the tool area sensor, k spindle is the interference coefficient of the vibration signal of the spindle vibration on the tool area, and the value range is [0, 1];

[0064]

[0065] is the signal after de-aliasing of the sensor in the workbench area, k base is the interference coefficient of the base vibration on the vibration signal of the workbench area.

[0066] It is understandable that for the vibration signals in the tool area and the worktable area, since the vibration signals in these areas are usually interfered by the vibrations in other areas, for example, the tool area signal may be affected by the coupling of the spindle vibration, and the worktable area signal may be superimposed by the resonance of the base area, it is necessary to perform de-aliasing to separate the independent vibration signals in each area. At the same time, since the signal may contain high-frequency noise, in order to ensure the purity of the signal and the accuracy of subsequent analysis, it is necessary to further perform denoising on the de-aliased signal.

[0067] For example, assuming that the vibration signal S in the tool area tool The spindle area signal S spindle The interference coefficient k spindle =0.4, and S tool =[5,10,15], S spindle =[1,2,3], then the signal after tool area dealiasing is The calculation result is Similarly, for the workbench area signal S workbench , if the base area signal S base Interference of the base, interference coefficient k base =0.3, and S workbench =[10,20,30],S base =[2,4,6], then the signal after dealiasing is The calculation result is The signals after de-aliasing and denoising can more realistically reflect the actual vibration characteristics of the tool area and the worktable area, laying the foundation for subsequent analysis.

[0068] Step S30: by performing a third multi-source independent signal data set S optimized Perform short-time Fourier transform to extract the time-frequency feature distribution The formula is:

[0069]

[0070] in, is the third multi-source independent signal dataset S optimized Any signal h(τ-t) is the Hamming window function with 128 sampling points; is the time-frequency distribution of the signal; t is the time center point of the current signal, f is the center value of the frequency component, and τ is the integral variable of time, which is used to traverse the entire time range of the signal;

[0071] From the time-frequency feature distribution Extract the key characteristics of machine tool vibration, including the main frequency component f main , vibration energy distribution E(f) and vibration amplitude dynamic range A range ;

[0072] is the third multi-source independent signal dataset S optimized middle Any signal constructs a feature vector, each of which contains the corresponding vibration main frequency component, vibration energy distribution and vibration amplitude dynamic range, and generates a feature vector set V optimized = {V spindle ,V tool ,V workbench ,V base}, where V spindle ,V tool ,V workbench ,V base They are the feature vector of the spindle area, the feature vector of the tool area, the feature vector of the worktable area and the feature vector of the base area;

[0073] It can be understood that the third multi-source independent signal data set of the signal processing logic is a signal after de-aliasing and denoising, and its signal quality is high, which can reflect the actual characteristics of the vibration signal in each area. By applying the Hamming window function to each signal and performing a short-time Fourier transform, a time-frequency distribution can be generated, from which three key features of the vibration signal can be extracted; the feature vectors generated by the signals in different areas correspond to the spindle area, tool area, worktable area and base area respectively, and these feature vectors together constitute a feature vector set. Through the feature vector set, the vibration characteristics of different areas of the machine tool can be fully described, providing input for the subsequent classification model.

[0074] It should be understood that compared with a single time domain or frequency domain analysis method, short-time Fourier transform can simultaneously obtain the time and frequency information of the vibration signal, which helps to identify the changing pattern of complex signals. For example, the high-frequency component of the vibration signal in the tool area can reflect the abnormality in the cutting process, while the low-frequency component of the signal in the worktable area may be related to the unstable clamping of the workpiece.

[0075] Step S40: Construct a machine tool vibration classification model, the classification model input is a feature vector set V optimized ,The classification model outputs the categories of vibration factors, including spindle unbalance vibration, tool cutting vibration, worktable clamping vibration and base environment interference vibration;

[0076] It should be noted that the output of the classification model is the category of vibration factors, including: spindle unbalance vibration, which is the unbalance vibration caused by the high-speed rotation of the spindle; tool cutting vibration, which is the high-frequency vibration caused by uneven cutting of the tool or tool wear during the processing; worktable clamping vibration, which is the low-frequency vibration caused by unstable workpiece clamping or abnormal worktable movement; base environment interference vibration, which is the vibration caused by low-frequency resonance of the external environment or unstable support of the machine tool base.

[0077] It is understandable that the classification model can effectively distinguish different types of vibration sources by analyzing the main frequency component, energy distribution and amplitude dynamic range in the input feature vector. For example, spindle unbalance vibration usually exhibits low-frequency and high-amplitude characteristics, while tool cutting vibration often exhibits high-frequency and high-energy characteristics. The classification model can combine these feature differences to achieve accurate classification of vibration sources; the dynamic convolutional neural network can dynamically adjust the weight of the convolution kernel according to the input feature vector, making the model more adaptable to different types of vibration signal characteristics, especially when the vibration signal is complex and the feature distribution is variable, which can significantly improve the classification accuracy.

[0078] It should be understood that the feature vector set V optimized = {V spindle ,V tool ,V workbench ,V base} is the core input of the classification model. Each feature vector contains the core features of the vibration signal in the corresponding area, which is the basis for the classification model to distinguish different vibration sources. For example: V spindle ={f main =200Hz,E(f)=0.45,A range =0.35} represents the vibration characteristics of the main axis area; V tool ={f main =10kHz,E(f)=0.65,A range =0.8} indicates the vibration characteristics of the tool area; the vibration factor categories output by the classification model can directly reflect the vibration sources of different areas of the machine tool, providing a basis for the analysis and optimization of the machine tool operation status. For example, if the classification results show that the tool cutting vibration is dominant, it can be judged that there may be problems of tool wear or cutting abnormalities during processing, thereby guiding tool replacement or processing parameter adjustment.

[0079] For example, suppose the set of feature vectors extracted from the third multi-source independent signal data set is: optimized = {V spindle ={20Hz,0.45,0.35},V tool ={10kHz,0.65,0.8},V workbench ={50Hz,0.3,0.4},V base={20Hz,0.2,0.25}}, where: the main axis region eigenvector It indicates that the main vibration frequency of the spindle is 200Hz, the energy is low, and the amplitude change is moderate, indicating that the spindle runs smoothly; the tool area feature vector It indicates that the main vibration frequency of the tool is 10kHz, with high energy and large amplitude variation, indicating that there may be cutting abnormalities or tool wear; based on the above feature vectors, the classification model outputs the following results: classification result = {spindle unbalance vibration: 20%, tool cutting vibration: 50%, worktable clamping vibration: 15%, base environmental interference vibration: 15%}. The results show that tool cutting vibration accounts for the main proportion.

[0080] Step S50: quantify the contribution ratio of the vibration source corresponding to the output vibration factor category and generate a detection report.

[0081] It should be noted that the test report includes the following contents: categories of vibration factors and their contribution ratios; key features (such as main frequency components, amplitude dynamic range, etc.); analysis of possible fault sources and improvement suggestions.

[0082] It should be noted that for each vibration factor i (such as spindle unbalance vibration, tool cutting vibration, etc.), its contribution ratio is i The calculation formula is:

[0083]

[0084] Among them, E i is the energy of the ith vibration factor, which is calculated from the output probability of the classification model and the energy distribution E(f) in the feature vector. total =∑ i E i is the total energy of the overall vibration signal.

[0085] It can be understood that the contribution ratio reflects the degree of influence of each vibration factor on the overall vibration of the target machine tool. For example, a high proportion of tool cutting vibration may mean that there is cutting abnormality or tool wear during the processing; a high proportion of base environmental interference may indicate that the external environment needs to be optimized.

[0086] In addition, the machine tool vibration detection system provided by the present invention adopts a machine tool vibration detection method in the above embodiment, which can solve the technical problem of machine tool vibration detection. Compared with the prior art, the beneficial effects of the machine tool vibration detection system provided by the present invention are the same as the beneficial effects of the machine tool vibration detection method provided by the above embodiment, and other technical features of the machine tool vibration detection system are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0087] The present invention provides a machine tool vibration detection device, please refer to Figure 2 , a machine tool vibration detection device includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a machine tool vibration detection method in the above-mentioned embodiment 1. A machine tool vibration detection device in the embodiment of the present invention may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. A machine tool vibration detection device is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention. A machine tool vibration detection device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM) 1004. In RAM 1004, various programs and data required for the operation of a machine tool vibration detection device are also stored. The processing device 1001, ROM 1002, and RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow a machine tool vibration detection device to communicate with other devices wirelessly or wired to exchange data. Although the figure shows a machine tool vibration detection device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or provided instead.

[0088] The present invention also provides a computer program product, including a computer program, wherein when the computer program is executed by a processor, the steps of the method for detecting vibration of a machine tool as described above are implemented. The computer program product provided by the present invention can solve the technical problem of detecting vibration of a machine tool. Compared with the prior art, the beneficial effects of the computer program product provided by the present invention are the same as the beneficial effects of the method for detecting vibration of a machine tool provided by the above embodiment, and will not be described in detail here.

[0089] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through a communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present invention are executed.

[0090] It should be understood that the various parts disclosed in the present invention can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0091] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for detecting vibration of a machine tool, characterized in that: Methods include: Step S10: installing multi-source sensors in the spindle area, tool area, worktable area and base area of ​​the target machine tool, including: a spindle area sensor, installed on the spindle bearing housing, used to record the unbalanced vibration of high-speed rotation; a tool area sensor, installed on the tool holder, used to collect cutting vibration and tool abnormality signals; a worktable area sensor, installed on the worktable base, used to capture vibration caused by worktable movement or unstable clamping of workpieces; a base area sensor, installed on the machine tool base, used to detect resonance interference transmitted to the machine tool by external vibration; The vibration signals of various areas of the target machine tool are collected in real time by multi-source sensors to form a first multi-source independent signal data set S collected ={S spindle ,S tool ,S workbench ,S base }, where S spindle is the vibration signal of the main axis area, S tool is the vibration signal of the tool area, S workbench is the vibration signal of the workbench area, S base is the vibration signal of the base area; Step S20: Vibration signal S for the tool area tool and the vibration signal S of the workbench area workbench Performing a de-aliasing process on the second multi-source independent signal data set to form a second multi-source independent signal data set, and performing a denoising process on the second multi-source independent signal data set to generate a third multi-source independent signal data set S optimized , in, is the denoised signal in the spindle region, is the signal after de-aliasing in the tool area, is the signal after de-aliasing in the workbench area, is the denoised signal of the base area; Step S30: by performing a third multi-source independent signal data set S optimized Perform short-time Fourier transform to extract the time-frequency feature distribution The formula is: in, is the third multi-source independent signal dataset S optimized Any signal h(τ-t) is the Hamming window function with 128 sampling points; is the time-frequency distribution of the signal; t is the time center point of the current signal, f is the center value of the frequency component, and τ is the integral variable of time, which is used to traverse the entire time range of the signal; From the time-frequency feature distribution Extract the key characteristics of machine tool vibration, including the main frequency component f main , vibration energy distribution E(f) and vibration amplitude dynamic range A range ; is the third multi-source independent signal dataset S optimized middle Any signal constructs a feature vector, each of which contains the corresponding vibration main frequency component, vibration energy distribution and vibration amplitude dynamic range, and generates a feature vector set V optimized = {V spindle ,V tool ,V worlbemch ,V nase }, where V spindle ,V tool ,V workbench ,V base They are the feature vector of the spindle area, the feature vector of the tool area, the feature vector of the worktable area and the feature vector of the base area; Step S40: Construct a machine tool vibration classification model, the classification model input is a feature vector set V optimized ,The classification model outputs the categories of vibration factors, including spindle unbalance vibration, tool cutting vibration, worktable clamping vibration and base environment interference vibration; Step S50: quantify the contribution ratio of the vibration source corresponding to the output vibration factor category and generate a detection report.

2. A method for detecting machine tool vibration as claimed in claim 1, characterized in that: In step S10, the multi-source sensor uses a three-axis acceleration sensor, model "ADXL356", with a range of ±40g, a sensitivity of 10mV / g, and a sampling frequency set to 20kHz.

3. A method for detecting machine tool vibration as claimed in claim 1, characterized in that: In step S20, the vibration signal S tool and the vibration signal S of the workbench area workbench To perform anti-aliasing, use the formula: is the signal after de-aliasing of the tool area sensor, k spindle is the interference coefficient of the vibration signal of the spindle vibration on the tool area, and the value range is [0, 1]; is the signal after de-aliasing of the sensor in the workbench area, k base is the interference coefficient of the base vibration on the vibration signal of the workbench area.

4. A method for detecting machine tool vibration as claimed in claim 1, characterized in that: In step S30, the vibration main frequency component f main The calculation formula is in, for The value of f at maximum; the calculation formula of vibration energy distribution E(f) is Vibration amplitude dynamic range A range The calculation formula is A range =A max -A min , where A max and A min are the maximum amplitude and minimum amplitude of the third multi-source independent signal respectively.

5. A method for detecting vibration of a machine tool as claimed in claim 1, characterized in that: In step S40, the step of outputting the vibration factor category by the classification model includes: The feature value F is extracted through the dynamic convolution layer conv (i,j), the calculation formula is: Among them, F conv (i, j) represents the feature value at position (i, j) in the output feature map after the convolution operation, W ′ is the dynamically adjusted convolution kernel weight, k is the size of the convolution kernel, C is the number of input channels, and b conv is the bias term, m, n represent the position offset of the convolution kernel in the vertical and horizontal directions respectively, and c is the input feature map channel; The probability distribution of the final classification is obtained through the fully connected layer and the Softmax function. The calculation formula is: P=Softmax(W fc ·F conv (i,j)+b fc ) Among them, P is the probability distribution of the vibration signal belonging to each category, W fc is the weight matrix of the fully connected layer, b fc is the bias parameter of the fully connected layer; Output the vibration factor category according to the value of P.

6. A method for detecting machine tool vibration as claimed in claim 5, characterized in that: Dynamic parameter W ′ The adjustment of is dynamically adjusted by inputting the feature vector, and its formula is: W ′ =W·σ(G) G=ReLU(W g ·V+b g ), Where W is the initial convolution kernel weight matrix; G is the dynamic adjustment factor generated by the input feature vector, which is calculated using the following formula: W ′ =W·σ(ReLU(W g ·V+b g )), where V is the input feature vector, including the main frequency component, amplitude dynamic range and spectrum energy; W g is a preset linear transformation matrix used to map the input feature vector to the space of dynamic adjustment factors; b g is the bias vector; ReLU is the activation function used to ensure that the output is non-negative; σ is the Sigmoid activation function, which is used to normalize the adjustment factor to the range of [0, 1].

7. A method for detecting machine tool vibration as claimed in claim 1, characterized in that: In step S50, the vibration source contribution ratio is calculated by the ratio of the spectrum energy of the vibration source.

8. A machine tool vibration detection system, characterized in that: The machine tool vibration detection system comprises: The vibration signal acquisition module is used to install multi-source sensors in the spindle area, tool area, worktable area and base area of ​​the target machine tool, including: a spindle area sensor installed on the spindle bearing housing to record the unbalanced vibration of high-speed rotation; a tool area sensor installed on the tool holder to collect cutting vibration and tool abnormality signals; a worktable area sensor installed on the worktable base to capture vibration caused by worktable movement or unstable clamping of workpieces; a base area sensor installed on the machine tool base to detect resonance interference transmitted to the machine tool by external vibration; The vibration signals of various areas of the target machine tool are collected in real time by multi-source sensors to form a first multi-source independent signal data set S collected ={S spindle ,S tool ,S workbench ,S base }, where S spindle is the vibration signal of the main axis area, S tool is the vibration signal of the tool area, S workbench is the vibration signal of the workbench area, S base is the vibration signal of the base area; Signal de-aliasing and denoising module for vibration signal S in the tool area tool and the vibration signal S of the workbench area workbench Performing a de-aliasing process on the second multi-source independent signal data set to form a second multi-source independent signal data set, and performing a denoising process on the second multi-source independent signal data set to generate a third multi-source independent signal data set S optimized , in, is the denoised signal in the spindle region, is the signal after de-aliasing in the tool area, is the signal after de-aliasing in the workbench area, is the denoised signal of the base area; The feature extraction module is used to extract the third multi-source independent signal data set S optimized Perform short-time Fourier transform to extract the time-frequency feature distribution The formula is: in, is the third multi-source independent signal dataset S optimized Any signal h(τ-t) is the Hamming window function with 128 sampling points; is the time-frequency distribution of the signal; t is the time center point of the current signal, f is the center value of the frequency component, and τ is the integral variable of time, which is used to traverse the entire time range of the signal; From the time-frequency feature distribution Extract the key characteristics of machine tool vibration, including the main frequency component f main , vibration energy distribution E(f) and vibration amplitude dynamic range A range ; is the third multi-source independent signal dataset S optimized middle Any signal constructs a feature vector, each of which contains the corresponding vibration main frequency component, vibration energy distribution and vibration amplitude dynamic range, and generates a feature vector set V optimized = {V spindle ,V tool ,V workbench ,V base }, where V spindle ,V tool ,V workbench ,V base They are the feature vector of the spindle area, the feature vector of the tool area, the feature vector of the worktable area and the feature vector of the base area; The vibration factor classification module is used to build a machine tool vibration classification model. The input of the classification model is the feature vector set V optimized ,The classification model outputs the categories of vibration factors, including spindle unbalance vibration, tool cutting vibration, worktable clamping vibration and base environment interference vibration; The test report output module is used to quantify the contribution ratio of the vibration source corresponding to the output vibration factor category and generate a test report.

9. A machine tool vibration detection device, characterized in that: The machine tool vibration detection device includes: a memory, a processor, and a machine tool vibration detection program stored in the memory and executable on the processor. When the machine tool vibration detection program is executed by the processor, the machine tool vibration detection method described in any one of claims 1 to 7 is implemented.

10. A computer program product, characterized in that The computer program product comprises a machine tool vibration detection program, and when the machine tool vibration detection program is executed by a processor, the machine tool vibration detection method described in any one of claims 1 to 7 is implemented.

Images