An intelligent construction equipment abnormality online detection method based on image recognition technology

Conventional intelligent construction equipment anomaly detection is based on the extraction and fusion of anomalous feature signals, which does not take into account the influence of noise signals on detection. Therefore, an online detection method of intelligent construction equipment anomaly based on image recognition technology is designed. Screening of intelligent construction equipment operation data abnormal mutation points, using a sliding window to traverse the historical data, labeling the abnormal mutation points, to provide the basic conditions for subsequent detection. Based on the image recognition technology to build intelligent construction equipment anomaly online detection model, according to the discriminant function and discriminant rules of the image recognition technology to detect the anomalies of the construction equipment, to avoid the problem of detection errors. The online detection of abnormal target integrity of construction equipment, combined with the operation data of the equipment, detects the dynamic changes of key components, so as to meet the demand for comprehensive abnormality detection of construction equipment. Using comparative experiments, it is verified that the anomaly detection results of this method are more accurate and can be applied in real life.