Tilt Angle Monitoring and Cable Controlling of Holding Pole via FEA and SR-GRU

The holding pole is a kind of special lifting machinery, which is widely used in the erection of power transmission tower. The holding pole belongs to the highrise structure, its stability problem is particularly prominent. Due to the high safety requirements, it is particularly important to conduct in-depth mechanical analysis before design, construction and application. For the holding pole, it has been adopted the balanced lifting as far as possible, however, the construction process will inevitably appear unbalanced lifting. Therefore, two working conditions for the holding pole, namely balanced lifting and unbalanced lifting, are discussed based on the Finite Element Analysis (FEA) in this paper. In addition, introducing the Sparse Residual (SR) learning method to accelerate the training of the applied Gated Recurrent Unit (GRU) network and avoid over-fitting, which can provide the advanced interfere for the unbalanced lifting through the real-time series data collected by accelerometer sensor. Finally, the experimental results from a project case verify the efficiency and robustness of proposed approaches.