Method Research Of Condition Monitoring And Fault Diagnosis For Tower Cranes

Tower cranes are the major equipment that are used widely in industrial and civil building, and their security issue has been widely public concern. It's hard to conclude the nature of accident when the accident occurred. Along with the development of AI, advanced intelligent of fault diagnosis technology will be also applied on tower crane. This text is based on QTZ80-type tower cranes to research its condition monitoring and fault diagnosis. Its major contents have:(1)Through detailed investigation for the current security situation of Tower crane and depth analysis for intelligent diagnosis technology, discuss the existing solutions of fault diagnosis problem on tower crane.(2)Torque is an important parameter of tower cranes. In this paper, through studying several common methods of the torque measurements, functional-link neural network method is gave. Establishing the nonlinear model of torque improves accuracy and stability of monitoring system for on line monitoring to provide a new approach.(3)Based on multivariate statistical process control theory, by studying of modeling theory of Principal Component Analysis and Dynamic Principal Component Analysis,find that DPCA can overcome the small perturbations which would be false positives, and DPCA is broung out. Selecting the main monitoring variables of tower crane as research object, through pre-historical data to establish main meta-model.(4) Applicating fault diagnosis methods based-PCA to fault detection and diagnosis of tower cranes. Using well-established principal component model to simulate two sets of failure data. Applying SPE and Hotelling to determine whether the system has broke down, using contribution graph to identify fault sources. The results showed: the fault diagnosis method is feasible effectively, and can predict early fault of tower crane.This paper established fault detection and diagnosis model of tower cranes successfully and provided a new method for line-measurement of torque, the results for tower tranes provide theoretical foundation, which have broad application prospect.