Sensorless Adaptive SMO With Mixed Order Generalized Integrator Based Improved QPLL PMSM Drive for LEV Application

The conventional sliding mode observer (SMO) with a low pass filter is constrained by several limitations due to discontinuous switching functions leading to degraded performance, particularly at low speeds. This paper introduces a novel adaptive SMO for permanent magnet synchronous motor (PMSM) drive, facilitating sensorless speed and position estimation. To reduce chattering effect, an adaptive gain-based sigmoid function is used as a switching function. In addition to this, harmonics contain, and DC offset from estimated back-EMF are reduced by employing mixed order generalized integrator (MOGI) as prefilter. To extract rotor speed and position and to mitigate harmonics arising from inverter nonlinearities, an improved quadrature phase locked loop (IQPLL) is utilized. This approach enhances QPLL estimation performance without compromising bandwidth or introducing phase delays typically associated with conventional QPLLs as well as achieving selective harmonics cancellation. Collective findings demonstrate effectiveness of proposed methodology in enhancing speed and position estimation accuracy while addressing inherent challenges associated with conventional techniques. Overall sensorless SPMSM performance is supported by obtained results.