The paper presents a method of processing vibration signals which was designed to detect damage to wheels of gearboxes for means of transport. This method uses entropy calculation, and multi-stage filtering calculated by means of digital filters and the Walsh-Hadamard transform to process signals. The presented method enables the extraction of vibration symptoms, which are symptoms of gear damage, from a complex vibration signal of a gearbox. The combination of multi-stage filtering and entropy enables the elimination of fast-changing vibration impulses, which interfere with the damage diagnosis process, and make it possible to obtain a synthetic signal that provides information about the state of the gearing. The paper demonstrates the usefulness of the developed method in the diagnosis of a gearbox in which two types of gearing damage were simulated: tooth chipping and damage to the working surface of the teeth. The research shows that the application of the proposed method of vibration of signal processing enables observation of the qualitative and quantitative changes in the entropy signal after denoising, which are unambiguous symptoms of the diagnosed damage.
Keywords: Walsh–Hadamard transform; condition monitoring; digital filtering; entropy; gearbox.