Metal material fatigue remote monitoring system based on cloud computing

Abstract

The invention discloses a metal material fatigue remote monitoring system based on cloud computing, and relates to the technical field of metal material performance monitoring, and the method comprises the steps: collecting surface images, stress states and environment information through a multi-source sensor, preprocessing the information through an edge computing module, and uploading the information to a cloud; the cloud computing platform fuses the data and trains a fatigue prediction model to realize accurate prediction of the residual life; and the remote monitoring module supports multi-terminal real-time checking and grading early warning. The system solves the problems of multi-source data splitting and early warning lag in traditional monitoring, and is suitable for metal material fatigue monitoring in the fields of aerospace, energy equipment and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of metal material performance monitoring, and particularly relates to a metal material fatigue remote monitoring system based on cloud computing. BACKGROUND

[0002] During long-term service, metal materials are prone to fatigue failure under the influence of factors such as cyclic load and environmental corrosion, and the fatigue process is closely related to surface defect evolution, stress concentration and environmental parameters. The traditional monitoring method relies on offline sampling and flaw detection, and has the following defects: strong hysteresis: periodic detection cannot capture real-time fatigue state, and it is difficult to intervene when the crack expands to the critical value; data isolation: single-point detection data lacks global correlation, and cannot build a material service cycle model; lack of intelligence: the existing system mostly uses threshold comparison method, which cannot dynamically adapt to material property changes and complex working conditions.

[0003] Although the existing technology realizes remote data transmission, it only uses simple threshold early warning and does not solve the problems of multi-dimensional analysis and dynamic prediction of fatigue data; although the laboratory high-throughput testing technology improves the data acquisition efficiency, it does not integrate cloud computing to realize remote monitoring and real-time intervention. Therefore, an intelligent monitoring system integrating the advantages of edge computing and cloud computing is urgently needed. SUMMARY

[0004] In view of the above deficiencies of the prior art, the present application provides a metal material fatigue remote monitoring system based on cloud computing.

[0005] To achieve the above-mentioned application purposes, the technical scheme adopted by the present application is: A metal material fatigue remote monitoring system based on cloud computing, comprising a data acquisition module, an edge computing module, a cloud computing platform module, a remote monitoring module and a user interface module; The data acquisition module synchronously acquires surface image information, stress state information and environmental information of the metal material, and transmits them to the edge computing module through 5G or industrial Ethernet; The edge computing module pre-processes the data and performs real-time analysis, and triggers local early warning; The cloud computing platform module stores and deeply analyzes the pre-processed data, and trains a fatigue prediction model; The remote monitoring module displays data and pushes early warning through Web and mobile terminal; The user interface module supports multi-role permission management and remote operation.

[0006] Further, the data acquisition module comprises a surface image acquisition unit, a stress state acquisition unit and an environment information acquisition unit, the surface image acquisition unit adopts an industrial camera or a laser profile sensor; the stress state acquisition unit adopts a fiber grating sensor or a strain gauge array; and the environment information acquisition unit comprises a temperature sensor and a humidity sensor.

[0007] Further, the edge computing module comprises an image preprocessing unit, a stress data preprocessing unit and an edge early warning unit. The image preprocessing unit is used for denoising, enhancing and feature extracting the surface image information. The stress data preprocessing unit is used for removing high-frequency noise in the stress signal and marking abnormal fluctuations. The edge early warning unit identifies the surface crack in real time through a CNN model, and when one or more of the following conditions occurs, i.e., the crack length exceeds a threshold, the stress is out of limit or the instantaneous strain mutates, triggers a local sound-light alarm and pushes a warning to a user mobile terminal through a remote monitoring module.

[0008] Further, the fatigue prediction model is constructed based on an LSTM neural network and iteratively optimized by a federated learning model, and the input parameters include a stress amplitude, a cycle number, an environment temperature and a material hardness, and the model outputs metal fatigue state data with a timestamp; The fatigue prediction model further combines a K-means classification unit, which classifies the fatigue warning levels according to the metal fatigue state data and outputs classification information.

[0009] The present application has the following advantages: The present application collects surface images, stress states and environment information through multi-source sensors, uploads them to the cloud after preprocessing by the edge computing module; the cloud computing platform fuses the data and trains the fatigue prediction model to realize accurate prediction of the remaining life; and the remote monitoring module supports real-time viewing and hierarchical early warning of multiple terminals. The present application solves the problems of multi-source data fragmentation and early warning lag in traditional monitoring, and is suitable for fatigue monitoring of metal materials in the fields of aerospace, energy equipment and the like. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 The figure is a schematic diagram of the system architecture of the present application. DETAILED DESCRIPTION The specific embodiments of the present application are described below to facilitate understanding of the present application by those skilled in the art, but it should be clear that the present application is not limited to the scope of the specific embodiments, and it is obvious to those skilled in the art that various changes are within the spirit and scope of the present application as defined in the appended claims, and all applications utilizing the concept of the present application are within the scope of protection.

[0011] The metal material fatigue remote monitoring system based on cloud computing comprises a data acquisition module, an edge computing module, a cloud computing platform module, a remote monitoring module and a user interface module. The data acquisition module synchronously acquires surface image information, stress state information and environmental information of the metal material, and transmits the information to the edge computing module through 5G or industrial Ethernet. The edge computing module pre-processes the data and performs real-time analysis, triggering local early warning. The cloud computing platform module performs fusion storage and deep analysis on the pre-processed data, and trains a fatigue prediction model. The remote monitoring module displays data and pushes early warning through Web and mobile terminal. The user interface module supports multi-role permission management and remote operation.

[0012] The data acquisition module comprises a surface image acquisition unit, a stress state acquisition unit and an environmental information acquisition unit. The surface image acquisition unit adopts an industrial camera or a laser profile sensor. The stress state acquisition unit adopts a fiber grating sensor or a strain gauge array. The environmental information acquisition unit comprises a temperature sensor and a humidity sensor.

[0013] The edge computing module comprises an image pre-processing unit, a stress data pre-processing unit and an edge early warning unit. The image pre-processing unit is used for denoising, enhancing and feature extraction of the surface image information. In specific implementation, Gaussian filtering is adopted to denoise the surface image, adjust the contrast, and extract the crack edge based on the Canny operator to obtain the crack length information. The stress data pre-processing unit is used for removing high-frequency noise in the stress signal and marking abnormal fluctuations. In specific implementation, sliding average filtering is adopted to remove noise, and Z-score method is adopted to identify and eliminate outliers.

[0014] The edge early warning unit identifies the surface crack in real time through a CNN model. In specific implementation, the CNN model adopts MobileNetV3, and the stress data is processed through Kalman filtering. When one or more of the following conditions occur: the crack length exceeds the threshold, the stress is out of limit, or the instantaneous strain mutates, local audible and visual alarms are triggered, and the user mobile terminal is pushed for early warning through the remote monitoring module.

[0015] The fatigue prediction model is constructed based on an LSTM neural network, and is iteratively optimized through federated learning. The input parameters include stress amplitude, cycle number, environmental temperature and material hardness. The model output is metal fatigue state data with a timestamp. The fatigue prediction model further incorporates a K-means classification unit that classifies the fatigue warning levels according to the metal fatigue state data and outputs the classification information.

Claims

1. A metal material fatigue remote monitoring system based on cloud computing, characterized in that: Including data acquisition module, edge computing module, cloud computing platform module, remote monitoring module and user interface module; The data acquisition module synchronously collects surface image information, stress state information and environmental information of the metal material, and transmits it to the edge computing module via 5G or industrial Ethernet; The edge computing module pre-processes the data and performs real-time analysis to trigger local warnings; The cloud computing platform module performs integrated storage and in-depth analysis on the pre-processed data to train the fatigue prediction model; The remote monitoring module displays data and pushes warnings through the Web and mobile terminals; The user interface module supports multi-role authority management and remote operation.

2. The metal material fatigue remote monitoring system based on cloud computing according to claim 1 is characterized in that: The data acquisition module includes a surface image acquisition unit, a stress state acquisition unit and an environmental information acquisition unit. The surface image acquisition unit adopts an industrial camera or a laser profile sensor; the stress state acquisition unit adopts a fiber Bragg grating sensor or a strain gauge array; and the environmental information acquisition unit includes a temperature sensor and a humidity sensor.

3. The metal material fatigue remote monitoring system based on cloud computing according to claim 1 is characterized in that: The edge computing module includes an image preprocessing unit, a stress data preprocessing unit, and an edge warning unit; The image preprocessing unit is used to perform denoising, enhancement and feature extraction on the surface image information; The stress data preprocessing unit is used to remove high-frequency noise from the stress signal and mark abnormal fluctuations; The edge warning unit uses the CNN model to identify surface cracks in real time. When one or more of the following situations occurs: the crack length exceeds the threshold, the stress exceeds the limit, or the instantaneous strain mutation occurs, it triggers a local sound and light alarm and pushes an early warning to the user's mobile terminal through the remote monitoring module.

4. The metal material fatigue remote monitoring system based on cloud computing according to claim 1 is characterized in that: The fatigue prediction model is built based on an LSTM neural network and uses a federated learning iterative optimization model. Input parameters include stress amplitude, number of cycles, ambient temperature, and material hardness. The model outputs metal fatigue status data with a timestamp. The fatigue prediction model is also combined with a K-means classification unit, which classifies the fatigue warning level according to the metal fatigue status data and outputs classification information.

Images