The diagnosis of a gearbox transmission system using electrical control parameters

Most of the techniques used to monitor and diagnose faults from machines are usually based on additional measurements which require high setup costs and installation difficulties. This paper focuses on developing a new sensorless method to monitor and diagnose different faults of a gearbox transmission system based on the parameters acquired from control systems. The control data, which are available in most of machines, including armature current, load set point, speed demand, motor current, torque feedback and speed feedback have been explored based on a gearbox test system. A non linear regression model is adopted to correlate the datasets to obtain residuals from the observed and the predicted control parameters. Subsequently a model based method is implemented to detect common faults such as lower oil levels and shaft misalignment in the gearbox system. The results confirm that it is possible to use existing control parameters for monitoring such faults.