A self-sensing framework for weak fault detection of planetary gearbox.

Abstract

Planetary gearbox fault detection has attracted wide attention due to the planetary gearbox's key role in modern electro-mechanic equipment. However, traditional fault detection technologies still heavily rely on additional sensors. The resulting enormous cost of sensors restricts the application of those technologies. Given this situation, a self-sensing fault detection framework to explore the weak fault impulses of the planetary gearbox is presented without additional sensors. In this framework, we first capture the preliminary signals from the servo control systems. Then, the hole control model of the motor driving planetary gearbox is constructed. After this step, the feasibility of fault detection for the planetary gearbox through the motor servo control signals is investigated. With the measured servo control signals, a multi-signal assisting adaptive time synchronous averaging method is first proposed to explore fault impulses. This method first introduces a periodic enhanced Gini to select optimal parameters adaptively. Finally, experiments on a weak fault of three components in the planetary gearbox are carried out separately, certifying our framework's validation of planetary gearbox fault detection. This framework hopes to provide a novel scheme for the weak fault self-sensing of planetary gearboxes.