Thesis (Ph.D)-Cairo University - Faculty of Engineering - Department of Electrical Power and Machines Engineering

Bibliography: p. 109-125.

With the increased dependence on induction motors (IMs) in the modern industry, the detection of incipient motor faults becomes an imperative requirement to reduce maintenance costs and avoid unscheduled shutdowns. Broken bar faults (BBFs) and bearing faults are around 60%of motor faults. These faults are developed from high thermal stresses, excessive forces, environmental stresses and high currents that occur in the motor cage. This thesis proposes three protection schemes to detect and diagnose BBFs and bearing faults. The first scheme is introduced to detect BBFs and estimate fault severity in IMs under start-up conditions. It includes three main stages, applying a powerful optimized S-transform to the current signal, extracting the LSH from the (t-f) domain using a proposed adaptive (t-f) filter, and estimating a proposed fault severity index based on the energy of RLSH. The second scheme provides a novel adaptive scheme to detect and diagnosis BBFs in IMs during steady-state conditions. It can detect BBFs in their incipient phases including non-adjacent faults under variable inertia, variable loading conditions, and in a noisy environment. The main idea is to monitor continuously the variation in phase angle of the main sideband frequency components by applying Fast Fourier Transform for only one phase of stator current. The third scheme is introduced for bearing faults detection and diagnosis under fixed and time-varying speed conditions. It utilizes the persistence spectrum for monitoring bearing health condition, as it provides some features related to bearing health and fault conditions. In addition, a multi-scale structural similarity index is used as a robust basis for bearing faults detection and classification without the need for training process or expert knowledge The proposed schemesareextensively validated using simulation tests and/or experimental data that provedtheir effectivenessto detect and diagnose BBFs and bearing faults