IDR - IIT Kharagpur

Reactive Power Based Model Reference Adaptive System For Sensorless Induction Motor Drive

Reactive Power Based Model Reference Adaptive System For Sensorless Induction Motor Drive

 

Owing To The Merits Of Mechanical Ruggedness, Requirement Of Less Space, Suitabil- Ity In Applications Like Hazardous Environment And In Submersible Drives Etc., The Research And Development Of Speed Sensorless Drive Have Great Significance. Elim- Ination Of The Speed Encoder Also Makes The Drive Cheaper. Different Techniques Have Been Proposed To Implement A Sensorless Drive With Varying Performance And Complexity. Of These, A Class Of Controllers Has Used The Theory Of Model Reference Adaptive System (Mras) With Some Success. So Far Flux, Back-Emf And Reactive Power-Based Mras Are Reported In Literature. The Performance Is Satisfactory At High/Medium Speed, However, The Behavior At Low Speed Is Inadequate. . Mras Involves Two Models (Reference And Adjustable) And An Adaptation Mech- Anism. Therefore, Performance Of Such System Is Highly Dependent On The Selection Of Models And The Type Of Adaptation Mechanism. Flux Based System Requires Pure Integration And Also Depends On The Stator Resistance. The Pure Integration May Be Replaced By A Low Pass Filter And The Performance Of Such Mras Is Reported Upto 2hz. The Back-Emf Based Mras Is An Alternative. The Pure Integration Is Eliminated But At The Cost Of Current-Derivatives. Selection Of Reactive Power As The Functional Candidate To Form The Mras Has Also Been Reported. The Sys- Tem Is Now Immune To The Variation Of Stator Resistance. However, The System Exhibits Instability In The Regenerative Modes And The Current And Flux Derivatives Are Difficult To Handle. This Thesis Is An Attempt To Enhance The Performance Of Reactive Power Based Mras. A Modified Reactive Power Based Mras Is Proposed That Shows Enough Performance At All Speeds Including Zero Speed. Several Important Issues E.G. Sta- Bility, Sensitivity To Parameter Variation Etc. Have Been Addressed. Extensive

Recent Submissions

Search DSpace


Advanced Search

Browse

My Account

Discover

RSS Feeds