CN116166997A - A method, system, device and medium for diagnosing service status of an intelligent spindle - Google Patents

Abstract

The invention provides a method, a system, equipment and a medium for diagnosing the service state of an intelligent main shaft, which comprise the steps of constructing a data acquisition platform to realize data acquisition of the intelligent main shaft, carrying out data enhancement on an original signal by an overlapping sampling method, converting the original signal into a two-dimensional time-frequency image by CWT, and reserving the original signal after data enhancement as input of a one-dimensional channel; dividing the original signal after data enhancement and the time-frequency pattern into a training set, a testing set and a verification set according to the interval of 7:2:1 respectively; inputting the training set into an improved two-channel DenseNet model for training; inputting the verification set into the improved model to perform super-parameter optimization through Bayesian optimization; inputting the test set into the trained model to obtain an intelligent main shaft final state diagnosis result; according to the invention, through collecting intelligent spindle data, analyzing the collected signals, extracting more detail features, and providing basis for subsequent spindle performance evaluation.

Description

A method, system, device and medium for diagnosing service status of an intelligent spindle

technical field

The invention belongs to the technical field of intelligent manufacturing, and in particular relates to a method, system, equipment and medium for diagnosing the service state of an intelligent spindle.

Background technique

Intelligent manufacturing is an important new market for the manufacturing industry. The core competitiveness of enterprise intelligent manufacturing comes from the digitization and networking of equipment, mining and analysis of equipment processing data, and the construction of intelligent workshops, etc.; CNC machine tools are important equipment in the workshop, and the spindle as It is particularly important to realize the intelligence of the core components of CNC machine tools; in order to meet the processing requirements of the modern manufacturing industry and realize the intelligence of the spindle system, its primary task is to solve the problem of spindle perception. In order to improve the accuracy of system perception, considering the single physical field Contingency and singleness, through multi-sensor information fusion technology, data can complement each other, increase the diversity and complexity of information, effectively reduce the uncertainty caused by single-sensor information, and improve the stability and reliability of data ;By adding sensors such as acceleration, displacement, and temperature to the spindle system, collecting multi-physics information such as vibration, displacement, and temperature field, and using certain methods to judge the processing status and health status of the intelligent spindle through these information, and in case of failure It is of great significance to improve the processing efficiency of intelligent spindles to realize fault diagnosis.

In the actual processing process, the interference of the spindle due to factors such as vibration, temperature rise, and local cracks will affect the processing accuracy and efficiency; therefore, it is particularly important to improve the performance of the spindle system. Many processing processes today are not carried out alone. An interdependent relationship. In this complex processing environment, once the spindle system cannot meet the actual processing requirements, it will directly affect the next production process, which will eventually lead to a reduction in product production efficiency and processing quality; the existing technology generally adopts Expert experience, time-frequency domain analysis, wavelet analysis and other technologies to realize the autonomous perception of the main axis are becoming more and more difficult. Traditional machine learning methods such as BP neural network, SVM and hidden Markov model are limited by their shallow structure. It is impossible to extract the characteristic information that can characterize the running state of the spindle from the huge and complex measured signal data.

Contents of the invention

Aiming at the problems existing in the prior art, the present invention provides an intelligent spindle service state diagnosis method, system, equipment and medium, which can extract more complete detailed features of the spindle state, and provide effective identification and classification for subsequent failure modes. in accordance with.

The present invention is achieved through the following technical solutions:

A method for diagnosing the service state of an intelligent main shaft, comprising the following steps:

a. Build a data acquisition platform to realize the data acquisition of the intelligent spindle, enhance the original signal by overlapping sampling method, and convert the original signal into a two-dimensional time-frequency image through CWT, and retain the original signal after data enhancement, as an input to a one-dimensional channel;

b. Divide the original signal after data enhancement and the time-frequency map sample into training set, test set and verification set according to the interval of 7:2:1;

c. Input the training set into the improved dual-channel DenseNet model for training;

d. Input the verification set into the improved model to optimize hyperparameters through Bayesian optimization;

e. Input the test set into the trained model to obtain the final state diagnosis result of the intelligent spindle.

Further, the step c inputs the training set into the improved dual-channel DenseNet model for training; the improved dual-channel DenseNet model based on feature fusion consists of 2-input layer, 2-convolution layer and maximum pooling layer, 2 -Three-level dense connection block, 2-dense connection block with ECA, 2-three transition layers, 2-BN layer+Conv layer+Maxpooling layer, 2-LSTM layer, 2-flattening layer, one Concatenation channel merge , a fully connected layer and an output layer.

Further, the step d inputs the verification set into the improved model to perform hyperparameter optimization through Bayesian optimization; optimize the hyperparameters within a given range, and the hyperparameters to be optimized mainly include learning rate and batch size , training rounds, the number and size of convolution kernels, and the number of neurons in the fully connected layer.

Further, the step e inputs the test set into the model that has been trained to obtain the final state diagnosis result of the intelligent spindle;

Among them, TP is a true positive sample; TN is a true negative sample; FP is a false positive sample; FN is a false negative sample.

Further, the improvement process of the dual-channel DenseNet model is:

Complete the extraction of local and global features of the original signal by improving the combination of DenseNet network and LSTM network;

Adjust the parameters of the network and suppress overfitting through a fully connected layer and a dropout layer;

The output features are normalized by the normalized exponential function - Softmax function, and all output values are converted into probabilities, and the sum of all probability values is 1, where the Softmax function formula is:

Among them, j=1,...,K, K refers to the number of categories of the specific classification.

Further, the ECA attention mechanism is added to the last level of the densely connected block of the 2-3 densely connected block; ECA is based on the SE module, and the SE uses the fully connected layer FC to learn the channel attention information and changes it to 1* 1 Convolutional learning channel attention information, the specific steps include:

S1: First input the feature map, whose dimension is H*W*C;

S2: Perform spatial feature compression on the input feature map. In the spatial dimension, use the global average pooling GAP to obtain a 1*1*C feature map;

S3: Carry out channel feature learning on the compressed feature map to realize: through 1*1 convolution, learn the importance between different channels, and the output dimension at this time is still 1*1*C;

S4: Combining channel attention, the feature map 1*1*C of channel attention and the original input feature map H*W*C are multiplied channel by channel, and the feature map with channel attention is output.

Further, select the Adam optimizer to optimize the model, and the Adam optimizer for the model is:

Among them, M is the number of categories; y _ic is a sign function (0 or 1), if the true category of sample i is equal to C, it is 1, otherwise it is 0; p _ic is the predicted probability that observed sample i belongs to category c.

An intelligent spindle service state diagnosis system, including:

The acquisition module is used to build a data acquisition platform to achieve data acquisition of the intelligent spindle. The original signal is enhanced by overlapping sampling, and the original signal is converted into a two-dimensional time-frequency image by CWT, and the enhanced data is retained. The original signal, as the input of the one-dimensional channel;

The division module is used to divide the original signal and time-frequency map samples after data enhancement into training set, test set and verification set according to the interval of 7:2:1 respectively;

The training module is used to input the training set into the improved two-channel DenseNet model for training;

The optimization module is used to input the verification set into the improved model to perform hyperparameter optimization through Bayesian optimization;

The diagnosis module is used to input the test set into the trained model to obtain the final state diagnosis result of the intelligent spindle.

A computer device, comprising a memory, a processor, and a computer program stored in the memory and operable on the processor, when the processor executes the computer program, an intelligent spindle service state diagnosis method is implemented step.

A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of a method for diagnosing the service state of an intelligent spindle are realized.

Compared with the prior art, the present invention has the following beneficial technical effects:

The invention provides a method, system, equipment and medium for diagnosing the service state of an intelligent spindle, including building a data acquisition platform to realize data collection of the intelligent spindle, performing data enhancement on the original signal by overlapping sampling, and converting the original signal through CWT It is a two-dimensional time-frequency image, and the original signal after data enhancement is retained as the input of the one-dimensional channel; the original signal after data enhancement and the time-frequency image sample are divided into training set, Test set and verification set; input the training set into the improved dual-channel DenseNet model for training; input the verification set into the improved model for hyperparameter optimization through Bayesian optimization; input the test set into the trained In the model, the final state diagnosis result of the intelligent spindle is obtained; the present invention analyzes the collected signal by collecting the data of the intelligent spindle, and uses the combination of the improved dense connection network and the long-short-term memory network, based on the way of dual-channel fusion , and more detailed features are extracted, which provides a basis for the subsequent performance evaluation of the spindle.

Description of drawings

Fig. 1 is a flow chart of a method for diagnosing the service state of an intelligent spindle according to the present invention;

Fig. 2 is the flow chart of data acquisition and control of the spindle of the present invention;

Fig. 3 is a densely connected network structure diagram of the present invention;

Fig. 4 is the improved 2D-DenseNet model structural diagram of the present invention;

Fig. 5 is a structural diagram of a dual channel model of the present invention;

Fig. 6 is a structural diagram of the ECA attention mechanism of the present invention.

Detailed ways

The present invention will be further described in detail below in conjunction with specific embodiments, which are explanations of the present invention rather than limitations.

In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only It is an embodiment of a part of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

It should be noted that the terms "first" and "second" in the description and claims of the present invention and the above drawings are used to distinguish similar objects, but not necessarily used to describe a specific sequence or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances such that the embodiments of the invention described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having", as well as any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product or device comprising a sequence of steps or elements is not necessarily limited to the expressly listed instead, may include other steps or elements not explicitly listed or inherent to the process, method, product or apparatus.

The present invention provides a method for diagnosing the service state of an intelligent spindle, as shown in Figure 1, comprising the following steps:

a. Build a data acquisition platform to realize the data acquisition of the intelligent spindle, enhance the original signal by overlapping sampling method, and convert the original signal into a two-dimensional time-frequency image through CWT, and retain the original signal after data enhancement, as an input to a one-dimensional channel;

b. Divide the original signal after data enhancement and the time-frequency map sample into training set, test set and verification set according to the interval of 7:2:1;

c. Input the training set into the improved dual-channel DenseNet model for training;

d. Input the verification set into the improved model to optimize the hyperparameters through Bayesian optimization, and optimize the hyperparameters of the model through global optimization to play a role in fine-tuning the model;

e. Input the test set into the trained model to obtain the final state diagnosis result of the intelligent spindle.

It should be noted that in the step a, the vibration signal and the displacement signal of the main shaft are collected by the acceleration sensor and the displacement sensor; the filter and the signal amplifier are selected to perform A/D conversion on the collected signal; the processed signal is processed by the data acquisition card. The data is stored and transmitted; the received original signal is data enhanced by overlapping sampling method, the original signal is retained, and time-frequency domain analysis is carried out by CWT; the step b combines the data-enhanced original signal with the time-frequency diagram The samples are divided into training set, verification set, and test set according to the interval of 7:2:1; the signal after data enhancement is analyzed in time domain, frequency domain, and time-frequency domain, and the original The time-frequency diagram of the signal can clearly and accurately represent the time-frequency distribution of vibration.

It should be further explained that this application first uses densely connected networks to adaptively extract feature information and reduce dimensionality, and then builds batch normalization layers, convolutional layers, and maximum pooling layers to further extract deep features as the input of the LSTM layer. Finally, the extraction of global features is realized through the long-term short-term memory network (LSTM); in the present invention, the improved DenseNet network and LSTM network are used to extract local and global features, which can avoid the rapid growth of training parameters while the model is deepening and cause training speed The phenomenon is greatly slowed down, fewer LSTM units increase modeling capabilities, and the extraction of intelligent spindle fault features is more comprehensive.

Preferably, as shown in Figure 5, the step c inputs the training set into the improved dual-channel DenseNet model for training; the improved dual-channel DenseNet model based on feature fusion consists of 2-input layer, 2-convolution layer and Maximum pooling layer, 2-three-level dense connection block, 2-dense connection block with ECA, 2-three transition layers, 2-BN layer+Conv layer+Maxpooling layer, 2-LSTM layer, 2-flattening layer, a Concatenation channel combination, a fully connected layer and an output layer;

As shown in Figure 3, it is a dense connection network structure diagram of the present invention. M1: adopts a three-level dense connection method, and the dense connection is mainly composed of two parts: Dense block dense block + Transition layer transition block; M2: Dense layer improvement: BN , ReLU and 1x1Conv, BN, ReLU and 3x3Conv; M3: Transition layer improvement: BN, ReLU, 1x1Conv, average pool; M4: Add ECA attention module after the last dense block.

Preferably, the step d inputs the verification set into the improved model and performs hyperparameter optimization through Bayesian optimization; optimizes the hyperparameters within a given range, and the hyperparameters to be optimized mainly include learning rate and batch size , training rounds, the number and size of convolution kernels, the number of neurons in the fully connected layer, etc., the Bayesian optimization algorithm has many advantages, the number of iterations is small, the convergence speed is fast, and it still has strong robustness especially for non-convex problems. Stickiness.

Preferably, said step e inputs the test set into the model that has been trained to obtain the final state diagnosis result of the intelligent spindle;

Among them, TP is a true positive sample; TN is a true negative sample; FP is a false positive sample; FN is a false negative sample.

As shown in Figure 4, the improvement process of the dual-channel DenseNet model is:

Complete the extraction of local and global features of the original signal by improving the combination of DenseNet network and LSTM network;

Adjust the parameters of the network and suppress overfitting through a fully connected layer and a dropout layer. The dropout layer is generally added to the fully connected layer to prevent overfitting and improve the generalization ability of the model. Dropout is only used when training the model. When evaluating the model No dropout is required; the output features are normalized by the normalized exponential function - Softmax function, and all output values are converted into probabilities (between 0 and 1), and all probability values add up to 1, where the Softmax function The formula is:

Among them, j=1,...,K, K refers to the number of categories of the specific classification.

By flattening the multi-dimensional data into one-dimensional data through the Flatten() function in the flattening layer, the flattening dimensionality reduction process is realized; the parameters of the network are adjusted and over-fitting is suppressed through a fully connected layer and a dropout layer, and the dropout layer is generally It is added to the fully connected layer to prevent overfitting and improve the generalization ability of the model. Dropout is only used when training the model, and dropout is not required for model evaluation. During model evaluation, the dropout layer passes all activation units;

As shown in Figure 6, the ECA attention mechanism is added to the last level of densely connected blocks of the 2-3 densely connected blocks; on the basis of the SE module, the ECA uses the fully connected layer FC to learn channel attention information in SE, and changes the To learn channel attention information for 1*1 convolution, the specific steps include:

S1: First input the feature map, whose dimension is H*W*C;

S2: Perform spatial feature compression on the input feature map. In the spatial dimension, use the global average pooling GAP to obtain a 1*1*C feature map;

S3: Carry out channel feature learning on the compressed feature map to realize: through 1*1 convolution, learn the importance between different channels, and the output dimension at this time is still 1*1*C;

S4: Combining channel attention, the feature map 1*1*C of channel attention and the original input feature map H*W*C are multiplied channel by channel, and the feature map with channel attention is output.

Preferably, the Adam optimizer is selected to optimize the model, and the Adam optimizer is to the model as:

Among them, M is the number of categories; y _ic is a sign function (0 or 1), if the true category of sample i is equal to C, it is 1, otherwise it is 0; p _ic is the predicted probability that observed sample i belongs to category c.

The present invention provides an intelligent spindle service state diagnosis system, including:

The acquisition module is used to build a data acquisition platform to achieve data acquisition of the intelligent spindle. The original signal is enhanced by overlapping sampling, and the original signal is converted into a two-dimensional time-frequency image by CWT, and the enhanced data is retained. The original signal, as the input of the one-dimensional channel;

The division module is used to divide the original signal and time-frequency map samples after data enhancement into training set, test set and verification set according to the interval of 7:2:1 respectively;

The training module is used to input the training set into the improved two-channel DenseNet model for training;

The optimization module is used to input the verification set into the improved model to perform hyperparameter optimization through Bayesian optimization;

The diagnosis module is used to input the test set into the trained model to obtain the final state diagnosis result of the intelligent spindle.

In yet another embodiment of the present invention, as shown in Figure 2, a hardware platform for a method for diagnosing the service state of an intelligent spindle includes: a three-way acceleration sensor, a thermocouple temperature sensor and a displacement sensor, a data acquisition card, a signal conditioning and storage device, Control module, piezoelectric actuator; this application first collects the original signal transmitted by the acceleration sensor, thermocouple temperature sensor, and displacement sensor; performs signal preprocessing through filters and amplifiers to remove noise and trend items; The obtained original signal is enhanced by overlapping sampling method, the original signal is retained, and the time-frequency domain analysis is performed through CWT to extract the time-frequency characteristics of the signal, which can better characterize the running state of the spindle; the model is completed by feature fusion The extraction of signal features; finally, the identification and classification of failure modes are completed by the softmax function.

In yet another embodiment of the present invention, a computer device is provided, the computer device includes a processor and a memory, the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the computer The program instructions stored in the storage medium. The processor may be a central processing unit (Central Processing Unit, CPU), or other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable GateArray, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, which are suitable for implementing one or more instructions, specifically for Loading and executing one or more instructions in the computer storage medium to realize the corresponding method flow or corresponding functions; the processor described in the embodiment of the present invention can be used for the operation of a method for diagnosing the service state of the intelligent spindle.

In yet another embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device for storing programs and data. . It can be understood that the computer-readable storage medium here may include a built-in storage medium in the computer device, and of course may also include an extended storage medium supported by the computer device. The computer-readable storage medium provides storage space, and the storage space stores the operating system of the terminal. Moreover, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions may be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one magnetic disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor, so as to realize the corresponding steps of the method for diagnosing the service state of the intelligent spindle in the above embodiments.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Accordingly, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it still The technical solutions described in the foregoing embodiments can be modified, or some or all of the technical features can be equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. The intelligent main shaft service state diagnosis method is characterized by comprising the following steps of:

a. constructing a data acquisition platform to realize data acquisition of the intelligent main shaft, carrying out data enhancement on an original signal by an overlapping sampling method, converting the original signal into a two-dimensional time-frequency image by CWT, and reserving the original signal after data enhancement as input of a one-dimensional channel;

b. dividing the original signal after data enhancement and the time-frequency pattern into a training set, a testing set and a verification set according to the interval of 7:2:1 respectively;

c. inputting the training set into an improved two-channel DenseNet model for training;

d. inputting the verification set into the improved model to perform super-parameter optimization through Bayesian optimization;

e. and inputting the test set into the trained model to obtain the final state diagnosis result of the intelligent spindle.

2. The intelligent spindle service state diagnosis method according to claim 1, wherein the step c is characterized in that a training set is input into an improved two-channel DenseNet model for training; the improved dual-channel DenseNet model based on feature fusion consists of a 2-input layer, a 2-convolution layer and a maximum pooling layer, a 2-three-level intensive connection block, a 2-intensive connection block with ECA, 2-three transition layers, a 2-BN layer+Conv layer+Maxpooling layer, a 2-LSTM layer, a 2-flattening layer, a localization channel combination, a full connection layer and an output layer.

3. The intelligent spindle service state diagnosis method according to claim 1, wherein the step d is characterized in that a verification set is input into an improved model to perform super-parameter optimization through Bayesian optimization; and optimizing the super-parameters in a given range, wherein the super-parameters to be optimized mainly comprise learning rate, batch size, training rounds, the number and size of convolution kernels and the number of neurons of a full-connection layer.

4. The method for diagnosing the service state of the intelligent spindle according to claim 1, wherein the step e is to input a test set into a trained model to obtain a final state diagnosis result of the intelligent spindle;

where TP is the true positive sample; TN is a true negative sample; FP is a false positive sample; FN is a false negative sample.

5. The intelligent spindle service state diagnosis method according to claim 1, wherein the improvement process of the dual-channel DenseNet model is as follows:

the extraction of local and global features of the original signal is completed by improving the combination mode of the DenseNet network and the LSTM network;

carrying out parameter adjustment and over-fitting inhibition on the network through a full connection layer and a dropout layer;

and (3) carrying out normalization processing on the output characteristics through a normalization exponential function-Softmax function, converting all output values into probabilities, wherein the sum of all probability values is 1, and the Softmax function formula is as follows:

where j=1, &.. K refers to the number of classes of a particular class.

6. The intelligent spindle service state diagnosis method according to claim 1, wherein an ECA attention mechanism is added to a final-stage dense connecting block of the 2-three-stage dense connecting block; based on the SE module, the ECA changes the FC learning channel attention information using the full connection layer in the SE into 1*1 convolution learning channel attention information, and the method specifically comprises the following steps:

s1: firstly, inputting a feature map, wherein the dimension of the feature map is H, W and C;

s2: performing spatial feature compression on the input feature map, and in the spatial dimension, using global average pooling GAP to obtain a 1x C feature map;

s3: and carrying out channel feature learning on the compressed feature map to realize: by 1*1 convolution, the importance among different channels is learned, and the output dimension is 1x C;

s4: channel attention combining, the channel attention feature map 1x C and the original input feature map H x W x C are multiplied channel by channel, and a feature map with channel attention is output.

7. The intelligent spindle service state diagnosis method according to claim 1, wherein an Adam optimizer is selected to optimize the model, and the Adam optimizer performs the following steps:

wherein M is the number of categories; y is _ic For a sign function (0 or 1), taking 1 if the true class of sample i is equal to C, otherwise taking 0; p is p _ic The predicted probability that sample i belongs to category c is observed.

8. An intelligent spindle service state diagnosis system, characterized in that, based on any one of the intelligent spindle service state diagnosis methods of claims 1-7, it comprises:

the acquisition module is used for constructing a data acquisition platform to realize data acquisition of the intelligent main shaft, carrying out data enhancement on the original signals by an overlapping sampling method, converting the original signals into two-dimensional time-frequency images by CWT, and reserving the original signals after data enhancement as input of a one-dimensional channel;

the dividing module is used for dividing the original signal after data enhancement and the time-frequency pattern into a training set, a testing set and a verification set according to the interval of 7:2:1 respectively;

the training module is used for inputting the training set into the improved two-channel DenseNet model for training;

the optimizing module is used for inputting the verification set into the improved model and carrying out super-parameter optimizing through Bayesian optimization;

and the diagnosis module is used for inputting the test set into the trained model to obtain the final state diagnosis result of the intelligent spindle.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of a method for diagnosing a service condition of an intelligent spindle according to any one of claims 1-7 when the computer program is executed by the processor.

10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the steps of a method for diagnosing a service condition of an intelligent spindle according to any one of claims 1-7.

Images