SDRSN-based multi-feature health factor fusion method
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A technology of health factors and fusion methods, applied in neural learning methods, testing of mechanical components, testing of machine/structural components, etc.
Active Publication Date: 2021-05-18
HARBIN INST OF TECH
View PDF13 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0005] The purpose of the present invention is to propose a multi-feature health factor fusion method based on SDRSN in view of the problem that the existing technology cannot guarantee that the activation value is transmitted between the various layers of the network in a normalized state and avoid over-fitting phenomenon
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
specific Embodiment approach 1
[0051] Specific implementation mode one: refer to figure 1 and figure 2 Specifically illustrate this embodiment, a kind of multi-feature health factor fusion method based on SDRSN described in this embodiment, it is characterized in that comprising:
[0052] Step 1: collecting the original vibration signal of the rotating machinery;
[0053] Step 2: Perform smoothing and denoising preprocessing on the original vibration signal of the rotating machinery, and then perform time domain, frequency domain and time-frequency domain feature extraction on the preprocessed original vibration signal of the rotating machinery, and construct the original feature set, Then normalize the signal in the original feature set;
[0054] Step 3: Use the normalized original feature set to filter and construct a sensitive feature set;
[0055] Step 4: Input the sensitive feature set into the SDRSN model for feature fusion training, input the data of the test set into the trained model, and obtai...
specific Embodiment approach 2
[0065] Embodiment 2: This embodiment is a further description of Embodiment 1. The difference between this embodiment and Embodiment 1 is that the final output of the model is expressed as:
[0066] x l+1 =x l +F(x l ,W l )
[0067] where x l Indicates the output feature A, F(x l ,W l ) represents the output feature B.
specific Embodiment approach 3
[0068] Embodiment 3: This embodiment is a further description of Embodiment 2. The difference between this embodiment and Embodiment 2 is that the convolutional layer output features are expressed as:
[0069] the y 1 =∑x*k+b
[0070] Among them, x represents the input feature, k represents the convolution kernel, and b represents the bias.
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more
PUM
Login to view more
Abstract
The invention discloses an SDRSN-based multi-feature health factor fusion method, relates to the technical field of fault prediction, and aims to solve the problem that a model in the prior art cannot reduce the influence of experience factors and remove redundant information. An SDRSN model can perform adaptive feature learning, discover interference features of an input sample according to an attention mechanism, and set the interference features to zero by using a soft threshold function, and therefore, the influence of interference factors on a feature mining effect is reduced. Compared with a traditional feature fusion method, the model can reduce the influence of experience factors and remove redundant information. The self-normalization thought is introduced into the SDRSN model, it can be guaranteed that an activation value is transmitted among layers of a network in a normalized state, the over-fitting phenomenon is avoided, features containing rich information are obtained, and therefore the health state of a rotating machine is better represented.
Description
technical field [0001] The invention relates to the technical field of fault prediction, in particular to a multi-feature health factor fusion method based on SDRSN. Background technique [0002] Due to its versatility, rotating machinery is currently widely used in various mechanical equipment and complex working environments. Once it is damaged, it will not only affect the normal use of the equipment, but may also cause huge economic losses and personal safety threats. Therefore, research on fault prediction methods for rotating machinery has always been an urgent need in the field of machinery health monitoring, and the construction of effective health factors is a prerequisite for accurate prediction of rotating machinery faults. [0003] Self-Normalizing Neural Networks (SNN) uses Scaled Exponential Linear Units (Scaled Exponential Linear Units, SELU) as the activation function, which can ensure that the activation value is transmitted between the layers of the network ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.