Gearbox variable working condition fault detection method based on normal sample aggregation contrast learning network

Abstract

A gearbox variable working condition fault detection method based on a normal sample aggregation contrast learning network comprises the following steps: using a vibration sensor to collect vibration signals of a gearbox running in different health states under the condition of rotating speed change, continuously performing offset transformation and synonym-preserving transformation on samples in a training set, and calculating the fault of the gearbox according to the offset transformation and synonym-preserving transformation; generating a virtual normal sample and virtual abnormal sample set; establishing a deep contrast learning network, calculating contrast loss based on normal sample aggregation and abnormal sample separation according to the generated virtual sample, and training the network by minimizing the contrast loss; inputting the memory sample set into the trained network, and determining a detection threshold by taking the minimum cosine distance between each sample and the remaining memory sample set as an abnormal score; and finally, calculating an abnormal score of the test sample, and judging the health state of the gearbox according to a threshold value. According to the invention, fault detection of the variable-working-condition gearbox is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical equipment state monitoring, and in particular to a gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network. Background Art

[0002] The health status monitoring and abnormal alarm of mechanical equipment are crucial to maintaining the safe operation of the equipment. Existing intelligent monitoring methods can accurately detect faults under a single working condition. However, the continuous fluctuation of working conditions under variable working conditions will lead to large differences in the collected signals, and the distribution of test samples and training samples will be offset, causing the monitoring model trained under fixed working conditions to produce false alarms or false alarms when facing vibration signals under variable working conditions. At the same time, the representations extracted by the existing detection models have poor fault discrimination and may be sensitive to working condition information such as speed, but not to the fault state. Therefore, the versatility of the fault detection model is limited in environments with variable speeds and frequent working condition switching. Self-supervised contrastive learning has excellent representation learning capabilities and provides a new idea for solving the above problems. However, the traditional contrastive learning objective excludes different normal samples, making it difficult to aggregate the representations of normal samples and generate representative normal representations under variable working conditions.

[0003] The above information disclosed in this Background section is only for enhancement of understanding of the background of the invention and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the invention

[0004] The present invention provides a gearbox variable operating condition fault detection method based on a normal sample aggregation contrast learning network, which continuously offsets and replaces the original signal to generate virtual samples, constructs a deep contrast learning framework, designs a contrast loss function with the goal of virtual normal sample aggregation and virtual abnormal sample separation, and trains the network. In the latent space, the minimum cosine distance between the sample to be tested and the memory sample is used as the abnormal score, and the detection threshold is determined based on the quantile of the abnormal score of the memory sample. Finally, the fault result is determined by comparing the abnormal score with the detection threshold.

[0005] A gearbox variable operating condition fault detection method based on a normal sample aggregation contrast learning network includes:

[0006] a. Use a vibration sensor to collect vibration signals of the gearbox running in different health states under the condition of changing speed, divide its time series into samples of fixed length and perform normalization processing, and construct a training sample set, a memory sample set and a test sample set respectively;

[0007] b. Continuously perform offset transformation and identity preservation transformation on the samples in the training set to generate a set of virtual normal samples and virtual abnormal samples;

[0008] c. Establish a deep contrastive learning network, which includes an encoder, a linear layer and a normalization layer, wherein the encoder is selected as a one-dimensional ResNet18, the linear layer and the normalization layer map the encoder output to a 128-dimensional hidden layer hypersphere space, and the activation function is set to a ReLU function. After the virtual normal samples and the virtual abnormal samples are input into the network, a hidden layer feature with a dimension of 128 is obtained, and a contrast loss based on normal sample aggregation and abnormal sample separation is calculated based on the hidden layer features. The network is trained using the gradient descent method, and the loss converges after 100 cycles of iterative training, and the model parameters are saved;

[0009] d. Input the memory sample set into the trained deep contrastive learning network, calculate the minimum cosine distance between each sample and the remaining sample set in the hidden layer hypersphere space as the anomaly score, and then determine the detection threshold;

[0010] e. Input the samples in the test sample set into the trained deep contrastive learning network, calculate the minimum cosine distance between the test samples in the test sample set and all the memory samples in the memory sample set as the anomaly score, and judge whether a fault exists based on the comparison between the anomaly score and the detection threshold.

[0011] In the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, in step a, the sample length after time series segmentation is 1024 data points, the amplitude is normalized to the range of [-1,1], 2700 samples are selected from the normal samples as the training sample set, 900 samples are selected as the memory sample set, and then 900 normal samples are selected to form a test sample set together with three types of fault types, 900 fault samples of each type.

[0012] In the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, in step b, the offset transformation is successfully achieved by randomly disrupting and reorganizing the arrangement of the signal. First, the original signal x is evenly divided into N segments. , disrupt and reorganize the order of each segment to form the offset transformed signal ,in .

[0013] In the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, in step b, the same transformation is performed twice on the original signal and the signal after the offset transformation, respectively. The original signal will generate a pair of virtual normal samples, and the signal after the offset transformation will generate a pair of virtual abnormal samples.

[0014] In the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, the same-preserving transformation includes the following steps:

[0015] Step 1: Add a value randomly sampled from a Gaussian distribution with a mean of 0 and a standard deviation of 0.01 to each sample point of the signal;

[0016] Step 2: Multiply each sample point by a value randomly sampled from a Gaussian distribution with a mean of 1 and a standard deviation of 0.01.

[0017] Step 3. Finally, resample the signal at a ratio of 0.8 and fill it with zeros to the original size.

[0018] In the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, in step c, the contrast loss is composed of a normal aggregation term and an abnormal separation term, and the overall loss form is:

[0019] ,

[0020] Where B is the batch size, and are the normal aggregation term and abnormal separation term of the i-th sample, respectively.

[0021] The normality aggregation term includes the contrast loss of each virtual normal sample, and the loss form is:

[0022]

[0023] in is a unit pulse function, which takes the value 1 if and only if the expression in the brackets holds, otherwise it takes the value 0. and are respectively a set of virtual normal samples and a set of virtual abnormal samples. and are two virtual samples generated by the i-th sample through identity-preserving transformation, It is the classic SimCLR contrast loss, and its loss form is:

[0024]

[0025] in, Respectively represent the input signal, the positive sample set of the input signal and the negative sample set of the input signal, is the set cardinality, is the hidden layer representation of the input, is a temperature hyperparameter used to adjust the loss. T represents the vector transpose operator.

[0026] The anomaly separation term includes the contrast loss of each virtual anomaly sample, and the loss form is:

[0027] .

[0028] in is a unit pulse function, which takes the value 1 if and only if the expression in the brackets holds, otherwise it takes the value 0. and are respectively a set of virtual normal samples and a set of virtual abnormal samples. and are two virtual samples generated by the i-th sample through identity-preserving transformation, It is the classic SimCLR contrast loss.

[0029] In the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, in step d, the abnormal score of the memory sample is the minimum cosine distance between the sample and the rest of the memory sample set, and is calculated as follows:

[0030] ,

[0031] Among them, x is the sample to be calculated, is a set of memory samples, represents the l2 norm of the vector, is the hidden layer representation of the input, Indicates the maximum value of the calculated results.

[0032] In the gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network, in step d, the detection threshold is the 95% quantile of the abnormal scores of all memory samples.

[0033] In the gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network, in step e, the calculation method of the test sample abnormality score is the same as the calculation method of the memory sample abnormality score.

[0034] In the gearbox variable operating condition fault detection method of the normal sample aggregation contrast learning network, for a continuously running variable speed gearbox, if the abnormality score calculated by the collected samples input into the deep contrast learning network exceeds the detection threshold for three consecutive times, it is considered that the variable speed gearbox has a fault.

[0035] Compared with the prior art, the present invention has the following advantages: the present invention fully improves the representation capability of vibration signals under different working conditions; generates virtual samples, and designs contrast loss with the goal of aggregating virtual normal samples and separating virtual abnormal samples, learns the ideal representation space for abnormality detection, and realizes the clustering of normal representations under different working conditions. Finally, the abnormality score is determined by distance calculation, which improves the accuracy and versatility of abnormality detection of gearboxes under variable working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] By reading the detailed description of the preferred specific embodiments below, various other advantages and benefits of the present invention will become clear to those of ordinary skill in the art. The drawings in the specification are only for the purpose of illustrating the preferred embodiments and are not considered to be limitations of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative work. Moreover, the same reference numerals are used to represent the same components throughout the drawings.

[0037] In the attached picture:

[0038] Figure 1 It is a flow chart of a gearbox variable operating condition fault detection method based on a normal sample aggregation contrast learning network provided by an embodiment of the present disclosure;

[0039] Figure 2 is a schematic diagram of three types of fault states of a planetary gearbox provided by an embodiment of the present disclosure;

[0040] FIG. 3 (a) to FIG. 3 (d) are schematic diagrams of vibration signals and speed changes of a planetary gearbox under different health states measured according to an embodiment of the present disclosure;

[0041] Figure 4 is a schematic diagram of performance comparison of a gearbox fault detection method provided by an embodiment of the present disclosure;

[0042] Figure 5 It is a schematic diagram of abnormal scores of samples in different health states under variable speed provided by an embodiment of the present disclosure.

[0043] The present invention is further explained below in conjunction with the accompanying drawings and embodiments. DETAILED DESCRIPTION

[0044] The specific embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the specific embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0045] It should be noted that certain words are used in the specification and claims to refer to specific components. Those skilled in the art should understand that technicians may use different nouns to refer to the same component. This specification and claims do not use the difference in nouns as a way to distinguish components, but use the difference in the functions of the components as the criterion for distinction. As mentioned throughout the specification and claims, "including" or "comprising" is an open term, so it should be interpreted as "including but not limited to". The subsequent description of the specification is a preferred embodiment of the present invention, but the description is based on the general principles of the specification and is not intended to limit the scope of the present invention. The scope of protection of the present invention shall be determined by the attached claims.

[0046] To facilitate understanding of the embodiments of the present invention, further explanation will be given below by taking specific embodiments as examples in conjunction with the accompanying drawings, and each of the accompanying drawings does not constitute a limitation on the embodiments of the present invention.

[0047] like Figures 1 to 5 As shown, the gearbox variable operating condition fault detection method of the normal sample aggregation contrast learning network includes the following steps:

[0048] a. Use a vibration sensor to collect vibration signals of the gearbox running in different health states under the condition of changing speed, divide its time series into samples of fixed length and perform normalization processing, and construct a training sample set, a memory sample set and a test sample set respectively;

[0049] b. Continuously perform offset transformation and identity preservation transformation on the samples in the training set to generate a set of virtual normal samples and virtual abnormal samples;

[0050] c. Establish a deep contrastive learning network, which includes an encoder, a linear layer and a normalization layer, wherein the encoder is selected as a one-dimensional ResNet18, the linear layer and the normalization layer map the encoder output to a 128-dimensional hidden layer hypersphere space, and the activation function is set to a ReLU function. After the virtual normal samples and the virtual abnormal samples are input into the network, a hidden layer feature with a dimension of 128 is obtained, and a contrast loss based on normal sample aggregation and abnormal sample separation is calculated based on the hidden layer features. The network is trained using the gradient descent method, and the loss converges after 100 cycles of iterative training, and the model parameters are saved; the deep contrastive learning network model training includes the following steps:

[0051] A one-dimensional ResNet18 network is used as the encoder, and virtual normal samples and virtual abnormal samples with a sample length of 1024 are input into the encoder to obtain features with a dimension of 512. Then, through a linear layer and a normalization layer with a dimension of 128, the activation function is set to the ReLU function to obtain the hidden layer features of virtual normal samples and virtual abnormal samples.

[0052] The overall contrast loss is calculated based on the hidden layer features, and the overall contrast loss is used as the objective function. The gradient descent method is used to train the network. After 100 cycles of iterative training, the loss converges and the model parameters are saved.

[0053] d. Input the memory sample set into the trained deep contrastive learning network, calculate the minimum cosine distance between each sample and the remaining sample set in the hidden layer hypersphere space as the anomaly score, and then determine the detection threshold;

[0054] e. Input the samples in the test sample set into the trained deep contrastive learning network, calculate the minimum cosine distance between the test samples in the test sample set and all the memory samples in the memory sample set as the anomaly score, and judge whether a fault exists based on the comparison between the anomaly score and the detection threshold.

[0055] In the preferred implementation of the gearbox variable operating condition fault detection method of the normal sample aggregation contrast learning network, in step a, the sample length after time series segmentation is 1024 data points, the amplitude is normalized to the range of [-1,1], 2700 samples are selected from the normal samples as the training sample set, 900 samples are selected as the memory sample set, and then 900 normal samples are selected to form a test sample set together with three types of fault types, 900 fault samples of each type.

[0056] In the preferred implementation of the gearbox variable operating condition fault detection method of the normal sample aggregation contrast learning network, in step b, the offset transformation is smoothly implemented by randomly disrupting and reorganizing the arrangement of the signal. First, the original signal x is evenly divided into N segments. , disrupt and reorganize the order of each segment to form the offset transformed signal ,in .

[0057] In a preferred implementation of the gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network, in step b, the same transformation is performed twice on the original signal and the signal after the offset transformation, respectively. The original signal will generate a pair of virtual normal samples, and the signal after the offset transformation will generate a pair of virtual abnormal samples.

[0058] In a preferred implementation of the gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network, the identity-preserving transformation includes the following steps:

[0059] Step 1: Add a value randomly sampled from a Gaussian distribution with a mean of 0 and a standard deviation of 0.01 to each sample point of the signal;

[0060] Step 2: Multiply each sample point by a value randomly sampled from a Gaussian distribution with a mean of 1 and a standard deviation of 0.01.

[0061] Step 3. Finally, resample the signal at a ratio of 0.8 and fill it with zeros to the original size.

[0062] In a preferred implementation of the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, in step c, the contrast loss consists of a normal aggregation term and an abnormal separation term, and the overall loss form is:

[0063] ,

[0064] Where B is the batch size, and are the normal aggregation term and abnormal separation term of the i-th sample, respectively.

[0065] The normality aggregation term includes the contrast loss of each virtual normal sample, and the loss form is:

[0066]

[0067] in is a unit pulse function, which takes the value 1 if and only if the expression in the brackets holds, otherwise it takes the value 0. and are respectively a set of virtual normal samples and a set of virtual abnormal samples. and are two virtual samples generated by the i-th sample through identity-preserving transformation, It is the classic SimCLR contrast loss, and its loss form is:

[0068]

[0069] in, Respectively represent the input signal, the positive sample set of the input signal and the negative sample set of the input signal, is the cardinality of the set, is the hidden layer representation of the input, is a temperature hyperparameter used to adjust the loss. T represents the vector transpose operator.

[0070] The anomaly separation term includes the contrast loss of each virtual anomaly sample, and the loss form is:

[0071] .

[0072] in is a unit pulse function, which takes the value 1 if and only if the expression in the brackets holds, otherwise it takes the value 0. and are respectively a set of virtual normal samples and a set of virtual abnormal samples. and are two virtual samples generated by the i-th sample through identity-preserving transformation, It is the classic SimCLR contrast loss.

[0073] In a preferred implementation of the gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network, in step d, the abnormal score of the memory sample is the minimum cosine distance between the sample and the rest of the memory sample set, and is calculated as follows:

[0074] ,

[0075] Among them, x is the sample to be calculated, is a set of memory samples, represents the l2 norm of the vector, is the hidden layer representation of the input, Indicates the maximum value of the calculated results.

[0076] In a preferred implementation of the gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network, in step d, the detection threshold is the 95% quantile of the abnormal scores of all memory samples.

[0077] In a preferred implementation of the gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network, in step e, the calculation method of the test sample abnormality score is the same as the calculation method of the memory sample abnormality score.

[0078] In a preferred implementation of the gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network, for a continuously running variable speed gearbox, if the abnormality score calculated by the collected samples input into the deep contrast learning network exceeds the detection threshold for three consecutive times, it is considered that the variable speed gearbox has a fault.

[0079] In one embodiment, a gearbox variable operating condition fault detection method using a normal sample aggregation contrast learning network includes:

[0080] In the first step, a vibration sensor is used to collect vibration signals of the gearbox under normal state (NS), broken tooth (BT), root crack (RC) and surface groove (SG) under the condition of speed change, with a sampling frequency of 10240Hz. The sample length is 1024. The signal time series is divided into several samples with a length of 1024 and the amplitude is normalized to the range of [-1,1]. 2700 samples are selected from the normal samples as the training sample set, 900 samples are selected as the memory sample set, and 900 normal samples are selected to form the test sample set together with three types of fault types and 900 fault samples of each type.

[0081] In the second step, the samples in the training set are continuously shifted and transformed to generate a set of virtual normal samples and virtual abnormal samples. The shift transformation is successfully achieved by randomly shuffling and reorganizing the arrangement of the signal. First, the original signal x is evenly divided into N segments , disrupt and reorganize the order of each segment to form the offset transformed signal ,in . Subsequently, two identity-preserving transformations are performed on the original signal and the shifted signal, respectively, including the following steps: adding a value randomly sampled from a Gaussian distribution with a mean of 0 and a standard deviation of 0.01 to each sample point of the signal; multiplying each sample point by a value randomly sampled from a Gaussian distribution with a mean of 1 and a standard deviation of 0.01; resampling the signal at a ratio of 0.8 and padding it to the original size. After continuously performing the shift and identity-preserving transformations, the original signal will generate a pair of virtual normal samples and a pair of virtual abnormal samples, which can be used as comparison examples for subsequent comparative learning.

[0082] In the third step, a deep contrastive learning network is established, which consists of an encoder, a linear layer and a normalization layer, wherein the encoder is selected as a one-dimensional ResNet18, and the linear layer and the normalization layer map the encoder output to a 128-dimensional hidden layer hypersphere space. Subsequently, based on the representation of the generated virtual normal samples and virtual abnormal samples in the hidden layer space, the contrast loss based on normal sample aggregation and abnormal sample separation is calculated, including the normal aggregation term and the abnormal separation term, and the overall form is:

[0083]

[0084] Where B is the batch size, and are the normal aggregation term and abnormal separation term of the i-th sample respectively. In the representation space, the ideal distribution is that normal samples are clustered together, while different types of faults are discretely distributed in different areas. Therefore, we hope that virtual normal samples are pulled closer, while virtual abnormal samples are repelled from normal samples and other types of abnormal samples. The normal aggregation term includes the contrast loss of each virtual normal sample, and the loss form is:

[0085]

[0086] The anomaly separation term includes the contrast loss of each virtual anomaly sample, and the loss form is:

[0087]

[0088] in It is a unit pulse function, and its value is 1 if and only if the expression in the brackets holds, otherwise it is 0. and are respectively a set of virtual normal samples and a set of virtual abnormal samples. and are two virtual samples generated by the i-th sample through identity-preserving transformation, It is the classic SimCLR contrast loss, and its loss form is:

[0089]

[0090] in, Respectively represent the input signal, the positive sample set of the input signal and the negative sample set of the input signal, is the cardinality of the set, is the hidden layer representation of the input, It is a temperature hyperparameter used to adjust the loss. T represents the vector transpose operator. Since the positive sample pairs are in the numerator and the negative sample pairs are in the denominator. Minimizing the normal aggregation term can shorten the distance between virtual normal samples, and minimizing the abnormal separation term can increase the distance between each type of virtual abnormal sample and the virtual normal sample and the remaining virtual abnormal samples, thereby forming a good abnormal detection representation space. After calculating the overall contrast loss, the contrast learning network is trained by minimizing the contrast loss. When the overall contrast loss converges, the network parameters are fixed.

[0091] In the fourth step, the memory sample set is input into the trained network, and the minimum cosine distance between each sample and the remaining sample set is calculated in the hidden space as the anomaly score, which is calculated as follows;

[0092]

[0093] Among them, x is the sample to be calculated, is a set of memory samples, represents the l2 norm of the vector, is the hidden layer representation of the input, Indicates the maximum value of the calculated result. Since all memory samples are normal samples, they follow the characteristic distribution of normal samples and have a small minimum cosine distance with the rest of the memory samples. Therefore, after the anomaly score calculation is completed, the detection threshold is selected as the 95% quantile of the anomaly score of all memory samples.

[0094] In the fifth step, the samples in the test set are input into the trained network, and the minimum cosine distance between the test sample and all memory samples is calculated as the anomaly score. The presence of a fault is determined by comparing the anomaly score with the detection threshold. In order to improve the robustness of the detection results, for a continuously running variable speed gearbox, if the calculated anomaly score exceeds the detection threshold for three consecutive times, the gearbox is considered to be faulty.

[0095] Figure 2Three types of faulty gears other than the normal state in the gearbox fault presetting experiment are shown, including broken teeth, tooth root cracks and surface grooves.

[0096] Figures 3 (a) to 3 (d) show schematic diagrams of the vibration signal of a planetary gearbox changing with speed under four healthy conditions: normal, broken teeth, tooth root cracks, and surface grooves. It can be clearly seen that the amplitude of the vibration signal changes with the speed. The higher the speed, the greater the vibration amplitude. Therefore, the signal is affected by both the speed and the healthy state. It is difficult to determine whether the gearbox is faulty only from the time domain signal. Variable operating condition fault detection requires the model to extract fault-related features in order to make accurate detection results.

[0097] Figure 4 It is a schematic diagram of the performance comparison of the gearbox fault detection method of the embodiment of the present invention. The evaluation indicators of the detection results include the precision rate (TPR), the precision rate (TPR), the precision rate (TPR) and the area under the receiver operating curve (AUROC). The experimental results are compared with two models, namely the autoencoder and the traditional contrastive learning method. At the same time, the experimental results show the means and standard deviations of the four detection indicators obtained by each model in five experiments, as shown in Table 1.

[0098]

[0099] In this embodiment, normal sample aggregation comparison learning shortens the distance between normal sample features and learns a more representative normal sample feature distribution under variable working conditions, thereby achieving fault detection. Figure 4 It can be seen from the results in Table 1 that the proposed normal sample aggregation contrast learning has achieved the best performance in all detection indicators, and the stability of multiple experimental results is good, which verifies the superiority of the method.

[0100] Figure 5 The diagram of the abnormal scores of samples in different health states under variable speed is shown. It can be seen that although the speed is changing, the threshold determined by the abnormal score of the memory sample can well separate the test normal samples from the test fault samples. The normal samples are basically below the threshold line, while the fault samples of various types are basically above the threshold line, indicating that the method can well detect gearbox faults under variable working conditions.

[0101] Although the embodiments of the present invention are described above in conjunction with the accompanying drawings, the present invention is not limited to the above specific embodiments and application fields, and the above specific embodiments are only illustrative and instructive, rather than restrictive. A person of ordinary skill in the art can also make many forms under the guidance of this specification and without departing from the scope of protection of the claims of the present invention, all of which belong to the protection of the present invention.

Claims

1. A gearbox variable operating condition fault detection method based on a normal sample aggregation contrast learning network, characterized in that: The steps include: a. Use a vibration sensor to collect vibration signals of the gearbox running in different health states under the condition of changing speed, divide its time series into samples of fixed length and perform normalization processing, and construct a training sample set, a memory sample set and a test sample set respectively; b. Continuously perform offset transformation and identity preservation transformation on the samples in the training set to generate a set of virtual normal samples and virtual abnormal samples; c. Establish a deep contrastive learning network, which includes an encoder, a linear layer and a normalization layer, wherein the encoder is selected as a one-dimensional ResNet18, the linear layer and the normalization layer map the encoder output to a 128-dimensional hidden layer hypersphere space, and the activation function is set to a ReLU function. After the virtual normal samples and the virtual abnormal samples are input into the network, a hidden layer feature with a dimension of 128 is obtained, and a contrast loss based on normal sample aggregation and abnormal sample separation is calculated based on the hidden layer features. The network is trained using the gradient descent method, and the loss converges after 100 cycles of iterative training, and the model parameters are saved; d. Input the memory sample set into the trained deep contrastive learning network, calculate the minimum cosine distance between each sample and the remaining sample set in the hidden layer hypersphere space as the anomaly score, and then determine the detection threshold; e. Input the samples in the test sample set into the trained deep contrastive learning network, calculate the minimum cosine distance between the test samples in the test sample set and all the memory samples in the memory sample set as the anomaly score, and judge whether a fault exists based on the comparison between the anomaly score and the detection threshold.

2. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1 is characterized in that: Preferably, in step a, the sample length after time series segmentation is 1024 data points, the amplitude is normalized to the range of [-1,1], 2700 samples are selected from the normal samples as the training sample set, 900 samples are selected as the memory sample set, and then 900 normal samples are selected to form a test sample set together with three types of fault types, 900 fault samples of each type.

3. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1 is characterized in that: In step b, the shift transformation is smoothly implemented by randomly shuffling and reorganizing the arrangement of the signal. First, the original signal x is evenly divided into N segments , disrupt and reorganize the order of each segment to form the offset transformed signal ,in .

4. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1 is characterized in that: In step b, the same transformation is performed twice on the original signal and the signal after the offset transformation, respectively. The original signal will generate a pair of virtual normal samples, and the signal after the offset transformation will generate a pair of virtual abnormal samples.

5. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 4 is characterized in that: The transformation includes the following steps: Step 1: Add a value randomly sampled from a Gaussian distribution with a mean of 0 and a standard deviation of 0.01 to each sample point of the signal; Step 2: Multiply each sample point by a value randomly sampled from a Gaussian distribution with a mean of 1 and a standard deviation of 0.

01. Step 3. Finally, resample the signal at a ratio of 0.8 and fill it with zeros to the original size.

6. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1 is characterized in that: In step c, the contrast loss consists of a normal aggregation term and an abnormal separation term, and the overall loss is: ； Where B is the batch size, and are the normal aggregation term and abnormal separation term of the i-th sample, respectively. The normality aggregation term includes the contrast loss of each virtual normal sample, and the loss form is: ； in is a unit pulse function, which takes the value 1 if and only if the expression in the brackets holds, otherwise it takes the value 0. and are respectively a set of virtual normal samples and a set of virtual abnormal samples. and are two virtual samples generated by the i-th sample through identity-preserving transformation, It is the classic SimCLR contrast loss, and its loss form is: ； in, Respectively represent the input signal, the positive sample set of the input signal and the negative sample set of the input signal, is the cardinality of the set, is the hidden layer representation of the input, is a temperature hyperparameter used to adjust the loss, T represents the vector transpose operator, The anomaly separation term includes the contrast loss of each virtual anomaly sample, and the loss form is: ； in is a unit pulse function, which takes the value 1 if and only if the expression in the brackets holds, otherwise it takes the value 0. and are respectively a set of virtual normal samples and a set of virtual abnormal samples. and are two virtual samples generated by the i-th sample through identity-preserving transformation, It is the classic SimCLR contrast loss.

7. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1 is characterized in that: In step d, the anomaly score of the memory sample is the minimum cosine distance between the sample and the rest of the memory sample set, calculated as follows: ； Among them, x is the sample to be calculated, is a set of memory samples, represents the l2 norm of the vector, is the hidden layer representation of the input, Indicates the maximum value of the calculated results.

8. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1 is characterized in that: In step d, the detection threshold is the 95th percentile of the abnormal scores of all memory samples.

9. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1 is characterized in that: In step e, the calculation method of the test sample anomaly score is the same as the calculation method of the memory sample anomaly score.

10. The gearbox variable operating condition fault detection method of a normal sample aggregation contrast learning network according to claim 1, characterized in that: For a continuously running variable speed gearbox, if the anomaly score calculated by inputting the collected samples into the deep contrast learning network exceeds the detection threshold for three consecutive times, the variable speed gearbox is considered to be faulty.

Images