Research and Implementation of Intelligent Warning Model for Drawworks Failure

Aiming at the problem that the fault signal noise components of the drawworks are complex due to the influence of base shaking, lateral wind load and external impacts such as lifting and lowering drill pipes and casing, making it difficult to achieve accurate fault early warning, a fault early warning method for the drawworks based on Iterative Singular Value Decomposition (ISVD) and Gauss mixture model (GMM) is proposed. ISVD is used to decompose the original signal and extract the effective components of the fault signal; the signal is reconstructed through Wavelet Packet Transform (WPT) to enhance the weak fault signal components; the GMM method is used to establish the statistical distribution, and combined with Kullback-Leibler Divergence (KL) for fault early warning. Through experimental verification, the intelligent early warning model combining GMM and KL can achieve an accuracy rate of 95.54% for drawworks fault early warning, effectively realizing the fault early warning of drawworks bearings.