Improved WD-Transformer Based Wire Rope Damage Detection Algorithm

In order to monitor the wear and tear of hoist ropes due to fatigue and other factors, and to ensure reliable hoist operation, a wire rope damage detection algorithm based on improved Wavelet Denoising (WD) and improved Transformer is proposed. Firstly, the wire rope monitoring data set is constructed, expanded and split by voltage sensor measurements. Subsequently, the wavelet denoising algorithm is improved by using a gradient descent automation design threshold to filter out voltage fluctuations due to wobbles and other factors. Next, the monitoring of the location, degree and type of wire rope defects and the determination of whether the wire rope meets the quality regulations are completed by constructing an improved Transformer computational model, including two parts, the Transformer prediction network and the multi-layer perceptron correction network. The experiments show that the above model has the advantage of high prediction accuracy compared with other similar models.