Application of an Adaptive Neuro Fuzzy Inference System for Low Speed Planetary Gearbox Vibration Control

Abstract

The gear pair assembly remains one of the major vibration sources in power transmission systems of mechanical rotating machinery. The gear vibration signature is often dominated by several high-level tonal peaks that occur at the fundamental gear mesh and its harmonics. The primary forcing function is produced by the gear transmission error excitation resulting from tooth profile errors, misalignment and elastic deformation. The excessive dynamic response generated can frequently lead to structural fatigue failure. Therefore, it is highly desirable to control the low-speed planetary gearbox rotational vibration acceleration levels. To achieve this goal, the present study demonstrates a method by which these vibrations produced due to the primary forcing function originating from the mesh points can be controlled. Hence, to deal with both the meshing stiffness generation and vibration transmissibility processes more effectively, the control system must be applied close to the gear's connection. This close proximity will allow the application of a single controller. Selection of the proper rule base depending upon the situation can be achieved by the use of an adaptive neuro-fuzzy inference system (ANFIS) controller as an integrated approach for purposes of control to yield excellent results, and this is the highlight of this paper. The results presented in the paper show that the outputs take less time to stabilize. Moreover, due to incorporation of the ANFIS controller with the plant, it is observed that the gearbox reaches the desired vibration very quickly in relatively shorter time.