Structure Fault Diagnosis Of Tower Crane Based On Time Series Analysis

Tower cranes is widely used in large-scale engineering and construction, mining, industry and other fields depend on the advantages of raised high, turning radius, lifting speed, etc.. However, because of its complex structure, high center of gravity, lifting capacity, agencies frequent starting and braking, it is also seen a big risk accident rate of machinery and equipment. Tower crane determines the safety and reliability. Analysis tower crane accident causes nearly half due to damage caused by metal structural damage.So, how to avoid and improve the reliability of its operation has become the focus of the industry attention.Damage identification method based on time series is a new research method in recent years which has been continuous developed. This paper introduces the method of structural damage identification of the tower crane. Provides a new way for tower crane health diagnostic study. The main tasks include:Time series theory was analyzed. Parameter estimation of time series model and the method of model order and model checking were studied. Judgment criterion of tower crane steel structure damage state recognition that is to extract time series model parameters as damage sensitive factor to do structural damage identification was proposed. Troubleshooting procedures for time series analysis were studied, and the tower crane steel structural damage criterion was stated. Extracted the time series model parameters as destroy sensitive factors for structural damage identification.The method of finite element modeling was studied as mini tower laboratory for the study. The micro-structural finite element model of the tower crane was established by the ANSYS software to conduce modal analysis of the overall structure of the tower crane. The dynamic model of the tower crane hoist off the ground was established against the dangerous working conditions of the tower crane lifting off the ground to solve the dynamic excitation loads. Took displacement information of important nodes obtained under different conditions tower crane model for transient analysis.Studied the AR model and residual variance analysis mothed. Selected the model residual variance as characteristic parameter to diagnose whether the structure damaged. Then designed a tower crane single limb experiments. Based the AR model with experimental data processed in accordance with the residual variance analysis to determine the state of the tower crane structure. Finally validated tower structure damage state identification based on AR model is a effectiveness method by different conditions.Proposed a new ARX model, taking the displacement signal of the tower different nodes as input, output data. Extract the damage characteristic parameters of the model. Combined with the statistical theory identify different states of tower structure. Then this method was validated by different conditions.The results show that the method of identifying the extent of the damage is effectively. Finally, introduced the principal component analysis method. Made a further research for tower crane damage diagnosis combined with the time series model.