Rotary machine life prediction method and system based on interactive learning convolutional network

Abstract

The invention discloses a rotating machine life prediction method and system based on an interactive learning convolutional network, and relates to the field of rotating machine life prediction, and the method comprises the steps: obtaining a rotating machine vibration signal, and extracting the shallow features of a time domain and a time-frequency domain; an ILCANet model framework is constructed; dividing the shallow features into a training sample set and a test sample set, and pasting labels to the training sample set to form a training label set; inputting the training sample set and the training label set into an ILCANet model for training, and inputting the test sample set into the ILCANet model after training is completed to obtain a prediction result; and comparing the difference between a prediction result and a true value, and verifying the precision and adaptability of the model. According to the method, different feature values of the vibration signals are extracted, weights are given through a feature attention mechanism, and the prediction precision is improved. Through odd-even segmentation and interactive learning, the problem that a convolutional network is insufficient in receptive field for long sequence data is solved.

Description

Technical Field

[0001] The present invention relates to the field of rotating machinery life prediction, and in particular to a rotating machinery life prediction method and system based on an interactive learning convolutional network. Background Art

[0002] As industrial development tends to be modernized, complex and sophisticated mechanical equipment is increasing. The development of deep learning technology has improved the remaining useful life (RUL) prediction performance of mechanical rotating parts. Traditional methods for predicting the remaining useful life of rolling bearings generally assume that a single vibration feature data is used to predict the life through a recurrent neural network. However, in many practical applications, this assumption does not hold true. For example, the complexity of the working environment means that a single feature cannot reflect the complex degradation process of the bearing, and the recurrent neural network cannot operate in parallel, making it difficult to predict the remaining useful life of the bearing in a timely manner.

[0003] Existing methods either use recurrent neural networks or use their variant networks, which do not take into account the time cost of prediction. The present invention intends to study a method for predicting the remaining service life of a bearing by relying on a convolutional neural network, so as to solve the problems of long training time of recurrent neural networks and prone to gradient explosion and gradient disappearance, thereby improving the remaining service life prediction system.

[0004] In view of the fact that the original sampling data of the rotating machinery vibration signal is too single, this paper attempts to solve this problem by extracting the shallow features of the original vibration signal in the time domain and time-frequency domain. This method can fully express the degradation information of the bearing. Summary of the invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the problem to be solved by the present invention is that the existing technical solutions do not take the time cost of prediction into consideration, and the original sampling data is too single.

[0007] To solve the above technical problems, the present invention provides the following technical solutions: a rotating machinery life prediction method based on an interactive learning convolutional network, which includes: obtaining a rotating machinery vibration signal, extracting shallow features in the time domain and time-frequency domain of the signal; constructing an ILCANet model framework; dividing the shallow features into a training sample set and a test sample set, and pasting labels on the training sample set to form a training label set, and the label pasting rule is that a group of vibration feature data corresponds to a type of life label; inputting the training sample set and the training label set into the ILCANet model for training, and after the training is completed, inputting the test sample set into the ILCANet model to obtain a prediction result; comparing the difference between the prediction result and the true value to verify the accuracy and adaptability of the model.

[0008] As a preferred solution of the rotating machinery life prediction method based on interactive learning convolutional network described in the present invention, wherein: the ILCANet model framework includes a one-dimensional convolution module, an interactive learning module, a feature attention module and a fully connected layer decoding module; the one-dimensional convolution module is composed of a copy filling layer, two one-dimensional convolution layers, a LeakyRelu activation layer, a Tanh activation layer and a Dropout layer, and its specific function is to map the input sample to a representation of the same dimension as the input.

[0009] As a preferred solution of the rotating machinery life prediction method based on interactive learning convolutional network described in the present invention, the interactive learning module includes performing point multiplication and addition and subtraction operations on the data after odd-even segmentation and the data processed by the one-dimensional convolution module, expressed as:

[0010]

[0011]

[0012] Among them, I odd Indicates odd data, I even represents even data, ⊙ represents Hadamard product, exp represents exponential function, c() represents data processed by one-dimensional convolution module, I o ' dd Represented as the final output odd data sequence, I e ' ven Represented as an even data sequence for the final output.

[0013] As a preferred solution of the rotating machinery life prediction method based on interactive learning convolutional network described in the present invention, the feature attention module includes receiving the data sequence output by the interactive learning module as the input x of the feature attention module, using global average pooling GAP to focus on the global view representation as:

[0014]

[0015] in, It is represented as the GAP output of the cth channel, j is represented as the jth group of data sequence, α is represented as the total number of sequences, and Mean is represented as the average operation; through the 1×16 convolution mapping function F1, it is x GAP Encode the dependency relationship with the input x, and get the feature matrix f expressed as,

[0016] f=Tanh(F1[cat(x,x GAP )])

[0017] Among them, cat represents the merged data; f is decomposed into x′ and others, and only x′ is retained. The feature matrix x′ contains the original input feature information and the global average pooling information. x′ is transformed into a variable with the same number of features as x through another convolution mapping function F2, which is expressed as,

[0018] g=Sigmoid[F2(f x′ )]

[0019] The fully connected layer decoding module is a decoder that maps the sequence to a RUL label.

[0020] As a preferred solution of the rotating machinery life prediction method based on interactive learning convolutional network described in the present invention, the training includes, when training the ILCANet model, minimizing the loss function MSELoss as the goal, and using an adaptive moment estimation method to adaptively adjust the learning rate of each parameter;

[0021] The loss function MSELoss is expressed as,

[0022]

[0023] Among them, y i Represents the i-th group of input sample set data, It is represented as the average value of the sample set data, and N is represented as the total number of data groups in the sample set.

[0024] As a preferred solution of the rotating machinery life prediction method based on interactive learning convolutional network of the present invention, wherein: the learning rate of adjusting each parameter is expressed as,

[0025]

[0026] m t =β1m t-1 +(1-β1)g t

[0027]

[0028]

[0029]

[0030]

[0031] Among them, g t Expressed as the gradient, Expressed as loss function, m t It is expressed as the first-order moment estimate of the gradient in momentum form at time t, v tIt is expressed as the second-order moment estimate of the gradient in momentum form at time t, and They are respectively the first-order and second-order moment estimates after bias correction, β1 and β2 are smoothing constants, and θ t It is represented as the parameter at time t, and η is represented as the learning rate.

[0032] As a preferred solution of the rotating machinery life prediction method based on interactive learning convolutional network described in the present invention, wherein: the verification model accuracy includes using mean absolute error MAE, root mean square error RMSE and scoring function SF to evaluate the prediction result and the true value, expressed as:

[0033]

[0034]

[0035]

[0036]

[0037] Among them, er t It is expressed as the difference between the predicted value and the true value at time t, n represents the total number of time steps, m represents the number of early time steps, m = n / 2, ω1 and ω2 represent the weight ratios of the early and late predictions of the bearing, respectively.

[0038] Another object of the present invention is to provide a rotating machinery life prediction system based on an interactive learning convolutional network, which can perform life assessment on rotating machinery.

[0039] To solve the above technical problems, the present invention provides the following technical solutions: a system for a rotating machinery life prediction method based on an interactive learning convolutional network, comprising: an acquisition module, a model building module, a data classification module, a training module and a verification module; the acquisition module is used to obtain the vibration signal of the rotating machinery and extract the shallow features in the time domain and the time-frequency domain of the signal; the model building module is used to construct an ILCANet model framework, which also includes a one-dimensional convolution module, an interactive learning module, a feature attention module and a fully connected layer decoding module; the data classification module is used to divide the shallow features into a training sample set and a test sample set, and to attach labels to the training sample set to form a training label set, and the label attaching rule is that a group of vibration feature data corresponds to a type of life label; the training module is used to input the training sample set and the training label set into the ILCANet model for training, and after the training is completed, the test sample set is input into the ILCANet model to obtain the prediction result; the verification module is used to compare the difference between the prediction result and the true value, and verify the accuracy and adaptability of the model.

[0040] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the rotating machinery life prediction method based on an interactive learning convolutional network are implemented as described above.

[0041] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the rotating machinery life prediction method based on an interactive learning convolutional network as described above.

[0042] The beneficial effects of the present invention are as follows: the present invention extracts different characteristic values ​​of the vibration signal and assigns weights through a feature attention mechanism, thereby improving prediction accuracy.

[0043] Through odd-even segmentation and interactive learning, the problem of insufficient receptive field of convolutional network for long sequence data is improved, the degree of feature extraction and life prediction performance in the life prediction process are significantly improved, and the robustness of life prediction is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. 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 creative work. Among them:

[0045] Figure 1 This is a flow chart of the rotating machinery life prediction method based on the interactive learning convolutional network in Example 1.

[0046] Figure 2 This is the ILCANet structure diagram of the rotating machinery life prediction method based on interactive learning convolutional network in Example 1.

[0047] Figure 3 This is a one-dimensional convolution module diagram of the rotating machinery life prediction method based on the interactive learning convolutional network in Example 1.

[0048] Figure 4 This is an interactive learning module diagram of the rotating machinery life prediction method based on the interactive learning convolutional network in Example 1.

[0049] Figure 5 This is a module diagram of the feature attention mechanism of the rotating machinery life prediction method based on the interactive learning convolutional network in Example 1. DETAILED DESCRIPTION

[0050] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0051] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0052] Example 1

[0053] Reference Figure 1-Figure 5 , which is the first embodiment of the present invention, and provides a rotating machinery life prediction method based on interactive learning convolutional network, including: Figure 1 As shown:

[0054] Step 1: Obtain the vibration signal of the rotating machinery and extract the shallow features of the signal in the time domain and time-frequency domain.

[0055] Step 2: If Figure 2 As shown in the figure, the ILCANet model framework is constructed. The ILCANet model framework includes a one-dimensional convolution module, an interactive learning module, a feature attention module, and a fully connected layer decoding module.

[0056] like Figure 3 As shown: The one-dimensional convolution module consists of a copy padding layer, two one-dimensional convolution layers, a LeakyRelu activation layer, a Tanh activation layer and a Dropout layer. Its specific function is to map the input sample to a representation of the same dimension as the input. The parameter is θ c .

[0057] The interactive learning module performs point multiplication and addition and subtraction operations on the data after odd-even segmentation and the data processed by the one-dimensional convolution module, which is expressed as:

[0058]

[0059]

[0060] Among them, I odd Indicates odd data, I even represents even data, ⊙ represents Hadamard product, exp represents exponential function, c() represents data processed by one-dimensional convolution module, I o ' dd Represented as the final output odd data sequence, I e ' ven Represented as an even data sequence for the final output.

[0061] like Figure 4 As shown: The feature attention module parameter is θ AThe input is the recombinant sequence output by the multi-layer interactive learning module, and the output is the weighted sequence data. Because there are too many data features, it is impossible to determine the contribution of each feature to the prediction of RUL. Therefore, the attention mechanism is used to capture the dependencies between data, pay more attention to important features, and prevent interference from useless information.

[0062] In it, global average pooling (GAP) is used to focus on the global view, because life prediction has a global receptive field, and GAP does not lose position information, so GAP is more suitable than global maximum pooling (GMP). The calculation formula of the c-th channel GAP is expressed as,

[0063]

[0064] in, It represents the GAP output of the cth channel, j represents the jth group of data sequence, α represents the total number of sequences, and Mean represents the average operation.

[0065] In addition, in order to find the position information, a 1×16 convolution mapping function F1 is used to obtain x GAP Encode the dependency relationship with the input x, and get the feature matrix f expressed as,

[0066] f=Tanh(F1[cat(x,x GAP )])

[0067] Among them, cat represents the merged data; then, f is decomposed into x′ and others, and only x′ is retained. The feature matrix x′ contains the original input feature information and the global average pooling information. x′ is transformed into a variable with the same number of features as x through another convolution mapping function F2, which is expressed as,

[0068] g=Sigmoid[F2(f x′ )]

[0069] like Figure 5 As shown: The parameters of the fully connected layer decoding module are θ F The fully connected layer is a decoder that maps the sequence to RUL labels.

[0070] Step 3: Divide the shallow features into a training sample set and a test sample set, and attach labels to the training sample set to form a training label set. The label attaching rule is that a set of vibration feature data corresponds to a type of life label.

[0071] Step 4: Input the training sample set and the training label set into the ILCANet model for training. After the training is completed, input the test sample set into the ILCANet model to obtain the prediction results.

[0072] When training the ILCANet model, the initialization strategy of the network model training is very important for building a deep architecture. In order to ensure the quality of the learned network parameters, the goal is to minimize the loss function MSELoss, and an adaptive moment estimation method is used to adaptively adjust the learning rate of each parameter.

[0073] The loss function MSELoss is expressed as,

[0074]

[0075] Among them, y i Represents the i-th group of input sample set data, It is represented as the average value of the sample set data, and N is represented as the total number of data groups in the sample set.

[0076] The learning rate for adjusting each parameter is expressed as,

[0077]

[0078] m t =β1m t-1 +(1-β1)g t

[0079]

[0080]

[0081]

[0082]

[0083] Among them, g t Expressed as the gradient, The loss function expression is consistent with the above loss function MSELoss, m t It is expressed as the first-order moment estimate of the gradient in momentum form at time t, v t It is expressed as the second-order moment estimate of the gradient in momentum form at time t, and They are respectively expressed as the bias-corrected first-order and second-order moment estimates, θ tIt is represented as the parameter at time t, η is the learning rate, β1 and β2 are smoothing constants, also known as attenuation coefficients. In machine learning, the default values ​​are usually β1 = 0.9 and β2 = 0.999, with a range of (0, 1). Its value is mainly based on the performance of the time series data in the data set. If the time series is relatively stable, the smoothing coefficient is smaller to reduce the correction amplitude so that the prediction model contains information from a longer time series; if the time series has a rapid and obvious tendency to change, the smoothing coefficient should be larger to improve the sensitivity of the prediction model. Experiments have proved that in most cases, β1 = 0.9 and β2 = 0.999 have the best optimization effect.

[0084] Step 5: Compare the predicted results with the true values ​​to verify the accuracy and adaptability of the model.

[0085] The mean absolute error MAE, root mean square error RMSE and scoring function SF are used to evaluate the prediction results and the true value, expressed as:

[0086]

[0087]

[0088]

[0089]

[0090] Among them, er t It is expressed as the difference between the predicted value and the true value at time t, n represents the total number of time steps, m represents the number of early time steps, m = n / 2, ω1 and ω2 represent the weight ratios of the early and late predictions of the bearing, respectively.

[0091] For bearing degradation, the prediction of late stage is more important than that of early stage, so a large amount of experimental data is analyzed with ω1 = 0.35 and ω2 = 0.65.

[0092] Example 2

[0093] The second embodiment of the present invention is different from the first embodiment in that the rotating machinery life prediction method based on the interactive learning convolutional network also includes, according to the interactive learning convolutional network bearing remaining service life prediction method based on the multi-feature attention mechanism proposed in this example, experiments are carried out on the phm2012 data set provided by the FEMTO-ST Institute. Through the method of the present invention: the experimental results of the interactive learning convolutional network bearing remaining service life prediction method based on the multi-feature attention mechanism are compared and analyzed with those of TCN-SA, TCN-RSA and TCN-RSCB, it is shown that the present invention has a significant improvement in MAE, RMSE and SF. It should be noted that the above-mentioned life prediction method is only illustrated by the experimental example of the above-mentioned benchmark data set. In actual applications, the above-mentioned method can be applied to different scenarios for experimental analysis as needed.

[0094] In this example, the database used in the experiment adopts the phm2012 rolling bearing full life test database collected by FEMTO-ST research through the PRONOSTIA test bench. The bearing degradation data mainly consists of two parts, namely vibration data and temperature data. The vibration sensor consists of two miniature accelerometers positioned at 90° to each other, one placed on the vertical axis and the other placed on the horizontal axis. The two accelerometers are placed radially on the outer ring of the bearing with a sampling frequency of 25.6KHz. The temperature sensor is a resistance temperature detector placed in a hole close to the outer bearing ring with a sampling frequency of 0.1Hz. The data set contains the full life data of different bearings under three working conditions, as shown in Table 1. Bearing is a bearing.

[0095] Table 1 Dataset details

[0096]

[0097] Since the vertical signal is less sensitive to bearing degradation, the horizontal vibration signals under the first two working conditions are selected for experimental research in the present invention, and the detailed description of the data set is shown in Table 1. Each bearing data in Table 1 is sampled at a sampling frequency of 25.6KHz, once every 10 seconds, and lasts for 0.1 seconds, that is, each sample contains 2560 data points.

[0098] In order to test the life prediction ability of the present invention, the training set and the test set were divided into cross-validation by the leave-one-out method. The model proposed in the present invention was compared with TCN-SA, TCN-RSA and TCN-RSCB, and the experimental results of the three evaluation index values ​​finally obtained are shown in Tables 2, 3 and 4.

[0099] Table 2 MAE of prediction results of different models

[0100]

[0101]

[0102] Table 3 RMSE of prediction results of different models

[0103]

[0104] Table 4 SF of prediction results of different models

[0105]

[0106]

[0107] From the prediction results, we can see that our method outperforms other methods on most bearings.

[0108] Example 3

[0109] The third embodiment of the present invention is different from the previous two embodiments in that: a system for a rotating machinery life prediction method based on an interactive learning convolutional network, comprising an acquisition module, a model building module, a data classification module, a training module and a verification module; the acquisition module is used to obtain the vibration signal of the rotating machinery and extract the shallow features in the time domain and the time-frequency domain of the signal; the model building module is used to build an ILCANet model framework, which also includes a one-dimensional convolution module, an interactive learning module, a feature attention module and a fully connected layer decoding module; the data classification module is used to divide the shallow features into a training sample set and a test sample set, and to attach labels to the training sample set to form a training label set, and the label attaching rule is that a group of vibration feature data corresponds to a type of life label; the training module is used to input the training sample set and the training label set into the ILCANet model for training, and after the training is completed, the test sample set is input into the ILCANet model to obtain the prediction result; the verification module is used to compare the difference between the prediction result and the true value, and verify the accuracy and adaptability of the model.

[0110] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0111] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0112] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0113] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0114] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A rotating machinery life prediction method based on interactive learning convolutional network, characterized by: include, Obtain the vibration signal of rotating machinery and extract the shallow features in the time domain and time-frequency domain of the signal; Construct the ILCANet model framework; The shallow features are divided into a training sample set and a test sample set, and labels are attached to the training sample set to form a training label set. The label attachment rule is that a set of vibration feature data corresponds to a type of life label. The training sample set and the training label set are input into the ILCANet model for training. After the training is completed, the test sample set is input into the ILCANet model to obtain the prediction results. Compare the differences between the predicted results and the actual values ​​to verify the accuracy and adaptability of the model.

2. The rotating machinery life prediction method based on interactive learning convolutional network according to claim 1, characterized in that: The ILCANet model framework includes a one-dimensional convolution module, an interactive learning module, a feature attention module, and a fully connected layer decoding module; The one-dimensional convolution module consists of a copy padding layer, two one-dimensional convolution layers, a LeakyRelu activation layer, a Tanh activation layer and a Dropout layer, and its specific function is to map the input samples into a representation of the same dimension as the input.

3. The rotating machinery life prediction method based on interactive learning convolutional network according to claim 2, characterized in that: The interactive learning module includes performing point multiplication and addition and subtraction operations on the data after odd-even segmentation and the data processed by the one-dimensional convolution module, which is expressed as: Among them, I odd Indicates odd data, I even represents even data, ⊙ represents Hadamard product, exp represents exponential function, c() represents data processed by one-dimensional convolution module, I o ' dd Represented as the final output odd data sequence, I e ' ven Represented as an even data sequence for the final output.

4. The rotating machinery life prediction method based on interactive learning convolutional network according to claim 3, characterized in that: The feature attention module includes receiving the data sequence output by the interactive learning module as the input x of the feature attention module, using global average pooling GAP to focus on the global view representation as: in, It is represented by the GAP output of the cth channel, j is represented by the jth data sequence, α is represented by the total number of sequences, and Mean is represented by the average operation; Through the 1×16 convolution mapping function F1, x GAP Encode the dependency relationship with the input x, and get the feature matrix f expressed as, f=Tanh(F1[cat(x,x GAP )]) Among them, cat represents the merged data; Decompose f into x′ and others, and only keep x′. The feature matrix x′ contains the original input feature information and the global average pooling information. Transform x′ into a variable with the same number of features as x through another convolution mapping function F2, expressed as: g=Sigmoid[F2(f x′ )] The fully connected layer decoding module is a decoder that maps the sequence to a RUL label.

5. The rotating machinery life prediction method based on interactive learning convolutional network according to claim 4, characterized in that: The training includes, when training the ILCANet model, minimizing the loss function MSELoss as a goal, and using an adaptive moment estimation method to adaptively adjust the learning rate of each parameter; The loss function MSELoss is expressed as, Among them, y i Represents the i-th group of input sample set data, It is represented as the average value of the sample set data, and N is represented as the total number of data groups in the sample set.

6. The rotating machinery life prediction method based on interactive learning convolutional network according to claim 5, characterized in that: The learning rate of adjusting each parameter is expressed as, m t =β1m t-1 +(1-β1)g t Among them, g t Expressed as the gradient, Expressed as loss function, m t It is expressed as the first-order moment estimate of the gradient in momentum form at time t, v t It is expressed as the second-order moment estimate of the gradient in momentum form at time t, and They are respectively the first-order and second-order moment estimates after bias correction, β1 and β2 are smoothing constants, and θ t It is represented as the parameter at time t, and η is represented as the learning rate.

7. The rotating machinery life prediction method based on interactive learning convolutional network according to claim 6, characterized in that: The accuracy of the verification model includes evaluating the prediction results and the true value using the mean absolute error MAE, the root mean square error RMSE and the scoring function SF, which is expressed as: Among them, er t It is expressed as the difference between the predicted value and the true value at time t, n represents the total number of time steps, m represents the number of early time steps, m = n / 2, ω1 and ω2 represent the weight ratios of the early and late predictions of the bearing, respectively.

8. A system using the rotating machinery life prediction method based on interactive learning convolutional network as claimed in any one of claims 1 to 7, characterized in that: It includes acquisition module, model building module, data classification module, training module and verification module; The acquisition module is used to obtain the vibration signal of the rotating machinery and extract the shallow features in the time domain and time frequency domain of the signal; The model building module is used to build the ILCANet model framework, which also includes a one-dimensional convolution module, an interactive learning module, a feature attention module and a fully connected layer decoding module; The data classification module is used to divide the shallow features into a training sample set and a test sample set, and to attach labels to the training sample set to form a training label set. The label attaching rule is that a set of vibration feature data corresponds to a type of life label; The training module is used to input the training sample set and the training label set into the ILCANet model for training, and after the training is completed, the test sample set is input into the ILCANet model to obtain the prediction result; The verification module is used to compare the difference between the predicted results and the true values, and to verify the accuracy and adaptability of the model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the rotating machinery life prediction method based on the interactive learning convolutional network described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the rotating machinery life prediction method based on an interactive learning convolutional network described in any one of claims 1 to 7 are implemented.

Images