CN-117485842-B - Method and system for monitoring attitude of wheel bucket cross beam of gate type bucket wheel machine in real time

CN117485842BCN 117485842 BCN117485842 BCN 117485842BCN-117485842-B

Abstract

The invention discloses a method and a system for monitoring the attitude of a wheel bucket cross beam of a gate type bucket wheel machine in real time. Firstly, acquiring a wheel bucket cross beam image of a wheel bucket cross beam acquired by a camera, then acquiring bending deformation amounts of the wheel bucket cross beam acquired by a displacement sensor at a plurality of preset time points in a preset time period, arranging the bending deformation amounts of the preset time points into bending deformation amount time sequence input vectors according to a time dimension, then carrying out image feature analysis on the wheel bucket cross beam image to obtain a wheel bucket cross beam feature map, then carrying out cross-modal element fusion feature analysis on the wheel bucket cross beam feature map and the bending deformation amount time sequence input vectors to obtain wheel bucket cross beam features fused with bending deformation time sequence features, and finally determining whether to generate wheel bucket cross beam safety precautions based on the wheel bucket cross beam features fused with the bending deformation time sequence features. The monitoring method and the system can improve the monitoring efficiency and accuracy.

Inventors

ZHAO XIA
YANG YANG
SUN XINJIA
TIAN HONGZHE
LIU PENGFEI
LIU CHANG
WANG YABIN
SU RUIZHI
ZHANG HAO
LIU XIANCHUN

Assignees

北京华能新锐控制技术有限公司

Dates

Publication Date: 20260505
Application Date: 20231103

Claims (6)

1. The method for monitoring the attitude of the wheel bucket cross beam of the gate type bucket wheel machine in real time is characterized by comprising the following steps: acquiring a wheel bucket cross beam image of a wheel bucket cross beam acquired by a camera; Acquiring bending deformation amounts of the wheel bucket cross beam acquired by the displacement sensor at a plurality of preset time points in a preset time period; Arranging the bending deformation amounts of the plurality of preset time points into bending deformation amount time sequence input vectors according to a time dimension; performing image feature analysis on the wheel bucket beam image to obtain a wheel bucket beam feature map; The wheel bucket beam characteristic diagram and the bending deformation time sequence input vector pass through a cross-modal element fusion module to obtain a wheel bucket beam characteristic diagram fused with the bending deformation time sequence characteristic as a wheel bucket beam characteristic fused with the bending deformation time sequence characteristic; the method comprises the steps of obtaining a bending deformation time sequence feature vector by passing the bending deformation time sequence input vector through a one-dimensional convolution layer of the cross-modal element fusion module, and obtaining a wheel bucket cross beam feature map fused with bending deformation time sequence features by taking the bending deformation time sequence feature vector as a channel weighting vector to carry out weighting processing along a channel dimension on the wheel bucket cross beam feature map; the method comprises the steps of carrying out feature distribution optimization on a wheel-bucket cross beam feature map of the fusion bending deformation time sequence feature to obtain a wheel-bucket cross beam feature map of the optimization bending deformation time sequence feature, and carrying out classification on the wheel-bucket cross beam feature map of the optimization fusion bending deformation time sequence feature through a classifier to obtain a classification result, wherein the classification result is used for showing whether the wheel-bucket cross beam safety precaution is generated or not, the wheel-bucket cross beam feature map of the fusion bending deformation time sequence feature is subjected to feature distribution optimization to obtain a wheel-bucket cross beam feature map of the optimization fusion bending deformation time sequence feature, and the method comprises the steps of calculating the global average value of feature matrixes of the wheel-bucket cross beam feature map along a channel dimension to obtain a wheel-bucket cross beam feature vector, carrying out optimization on the bending deformation time sequence feature vector by the wheel-bucket cross beam feature vector to obtain an optimization bending deformation time sequence feature vector, and carrying out weighting treatment on the wheel-bucket cross beam feature map along the channel dimension by the optimization bending deformation time sequence feature vector to obtain the wheel-bucket cross beam feature map of the optimization fusion bending deformation time sequence feature.
2. The method for monitoring the attitude of a wheel and bucket beam of a gate type bucket wheel machine in real time according to claim 1, wherein the image feature analysis is performed on the image of the wheel and bucket beam to obtain a wheel and bucket beam feature map, and the method comprises the following steps: And the wheel bucket beam image is passed through a wheel bucket beam posture feature extractor based on a convolutional neural network model to obtain the wheel bucket beam feature map.
3. The method for monitoring the attitude of the wheel and bucket beam of the gate-type bucket wheel machine in real time according to claim 2, wherein the wheel and bucket beam feature map which is optimized and fused with the bending deformation time sequence features is passed through a classifier to obtain a classification result, and the classification result is used for indicating whether the wheel and bucket beam safety precaution is generated or not, and the method comprises the following steps: expanding the wheel bucket cross beam characteristic diagram which is optimized and fused with the bending deformation time sequence characteristic into an optimized classification characteristic vector according to a row vector or a column vector; Fully-concatenated coding the optimized classification feature vector using a fully-concatenated layer of the classifier to obtain a coded classification feature vector, and And inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
4. A door-type bucket-wheel machine-wheel-bucket-beam attitude real-time monitoring system for performing the door-type bucket-wheel machine-wheel-bucket-beam attitude real-time monitoring method according to any one of claims 1 to 3, characterized by comprising: The image acquisition module is used for acquiring a wheel bucket cross beam image of the wheel bucket cross beam acquired by the camera; The bending deformation amount acquisition module is used for acquiring bending deformation amounts of the wheel bucket cross beam acquired by the displacement sensor at a plurality of preset time points in a preset time period; the vectorization module is used for arranging the bending deformation amounts of the plurality of preset time points into bending deformation amount time sequence input vectors according to the time dimension; The image feature analysis module is used for carrying out image feature analysis on the wheel bucket beam image to obtain a wheel bucket beam feature map; The cross-modal element fusion characteristic analysis module is used for carrying out cross-modal element fusion characteristic analysis on the wheel bucket beam characteristic graph and the bending deformation time sequence input vector to obtain wheel bucket beam characteristics fusing bending deformation time sequence characteristics, and And the safety analysis module is used for determining whether the wheel bucket cross beam safety early warning is generated or not based on the wheel bucket cross beam characteristics fused with the bending deformation time sequence characteristics.
5. The system for monitoring the attitude of a wheel and bucket beam of a gate type bucket wheel machine in real time according to claim 4, wherein the image feature analysis module is configured to: And the wheel bucket beam image is passed through a wheel bucket beam posture feature extractor based on a convolutional neural network model to obtain the wheel bucket beam feature map.
6. The real-time monitoring system for the attitude of a wheel and cross beam of a portal bucket wheel machine according to claim 5, wherein the cross-modal meta-fusion feature analysis module is configured to: And the wheel bucket cross beam characteristic diagram and the bending deformation time sequence input vector pass through a cross-modal element fusion module to obtain the wheel bucket cross beam characteristic diagram fused with the bending deformation time sequence characteristic as the wheel bucket cross beam characteristic fused with the bending deformation time sequence characteristic.

Description

Method and system for monitoring attitude of wheel bucket cross beam of gate type bucket wheel machine in real time Technical Field The application relates to the field of intelligent monitoring, in particular to a method and a system for monitoring the attitude of a wheel bucket cross beam of a door type bucket wheel machine in real time. Background The gate type bucket wheel machine is a large mechanical equipment for loading and unloading bulk materials (such as coal, ore, etc.), which is composed of a large gate type structure and a bucket wheel mechanism hung on the gate type structure, and is widely applied to the fields of thermal power plants, etc. The wheel bucket cross beam of the bucket wheel machine is a key component bearing load and bending moment, and the safety of the wheel bucket cross beam is important for guaranteeing the smooth operation of loading and unloading. However, due to long-term working environment and load changes, the bucket cross beam may be deformed, damaged by fatigue cracks and the like, and the safety and reliability of the door bucket turbine are affected. Therefore, the attitude and deformation condition of the wheel bucket cross beam are monitored in real time, abnormal conditions are found in time and early warning is carried out, and the method is an important measure for guaranteeing the normal and safe operation of the door type bucket wheel machine. However, the traditional bucket wheel and wheel bucket beam monitoring scheme mainly relies on professionals to carry out regular inspection and monitoring, and the mode needs to spend a large amount of manpower and time cost, and the monitoring efficiency and accuracy are lower. In addition, the traditional method can only acquire the state information of the wheel bucket cross beam during regular inspection, and cannot monitor the attitude and deformation condition of the wheel bucket cross beam in real time, so that some potential problems or abnormal conditions can be missed during the inspection process, and once the abnormality occurs, the processing time can be delayed, and the safety risk is increased. Therefore, a real-time monitoring scheme for the attitude of a wheel bucket beam of a gate type bucket wheel machine is desired. Disclosure of Invention The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides a method and a system for monitoring the attitude of a wheel and bucket beam of a gate type bucket wheel machine in real time, which can realize the real-time monitoring of the attitude and deformation condition of the wheel and bucket beam of the gate type bucket wheel machine, improve the monitoring efficiency and accuracy, help to discover the abnormal condition of the wheel and bucket beam in time and take corresponding measures, and ensure the safe operation of the bucket wheel machine. According to one aspect of the application, a method for monitoring the attitude of a bucket cross beam of a door type bucket wheel machine in real time is provided, which comprises the following steps: acquiring a wheel bucket cross beam image of a wheel bucket cross beam acquired by a camera; Acquiring bending deformation amounts of the wheel bucket cross beam acquired by the displacement sensor at a plurality of preset time points in a preset time period; Arranging the bending deformation amounts of the plurality of preset time points into bending deformation amount time sequence input vectors according to a time dimension; performing image feature analysis on the wheel bucket beam image to obtain a wheel bucket beam feature map; Performing cross-modal element fusion feature analysis on the wheel bucket cross beam feature map and the bending deformation time sequence input vector to obtain wheel bucket cross beam features fusing bending deformation time sequence features, and And determining whether to generate the wheel bucket cross beam safety early warning based on the wheel bucket cross beam characteristics fused with the bending deformation time sequence characteristics. According to another aspect of the present application, there is provided a system for monitoring the attitude of a wheel cross beam of a gate type bucket wheel machine in real time, comprising: The image acquisition module is used for acquiring a wheel bucket cross beam image of the wheel bucket cross beam acquired by the camera; The bending deformation amount acquisition module is used for acquiring bending deformation amounts of the wheel bucket cross beam acquired by the displacement sensor at a plurality of preset time points in a preset time period; the vectorization module is used for arranging the bending deformation amounts of the plurality of preset time points into bending deformation amount time sequence input vectors according to the time dimension; The image feature analysis module is used for carrying out image feature analysis on the wheel bucket beam image