Adaptive neuro fuzzy inference system based decoupled control for neutral point clamped multi level inverter fed induction motor drive

The presence of an integrator in a reference model rotor flux-based adaptive system (RF-MRAS) and non-linearity the inverter output voltage degrade speed response sensorless operation electric drive terms DC drift, initial value issues, inaccurate acquisition. To improve response, compensating component is supplemented by amending integrator. coalition drift offset voltages, reduces issues. During low-speed operation, acquisition distorts stator critically, it becomes considerable when machine low. Implementing three-level neutral point clamped speed-sensorless decoupled control induction motor improves performance with superior quality voltage. Further, improved replacing proportional-integral (PI) controller adaption mechanism RF-MRAS neuro-fuzzy inference (ANFIS) controller. A prototype (3L-NPC)-fed fabricated laboratory, its for RF-MRAS, modified RFMRAS, RFMRAS using ANFIS are compared different benchmark tests.