MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202631107807 A) filed by Indian Institute Of Technology Indian School Of Mines Dhanbad on September 08, 2026, for Method And System For Gearbox Condition Diagnosis Using Multi Domain Features And Cnn-Lstm-Gru Model.
Inventors include Dr. Sachin Kumar Singh; and Dr. Chandan Pandey.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: A method for diagnosing a gearbox condition is disclosed. Vibration and rotational signals are acquired from sensors of an input unit (100) and segmented into analysis windows using a signal segmentation module (226). Frequency domain features are extracted using a Fast Fourier Transform module (222), and time–frequency features are extracted using a Continuous Wavelet Transform module (224). The features are combined to form a unified feature representation. A convolutional neural network–long short term memory–gated recurrent unit model (228) processes the unified representation by applying convolutional layers to learn spatial patterns and recurrent layers to learn temporal dependencies, generating a fault classification output (304). A gearbox fault condition label is determined based on the classification output, enabling accurate diagnosis under varying operating conditions and limited target domain data.
Disclaimer: Curated by HT Syndication.
