Please use this identifier to cite or link to this item: https://dspace.upt.ro/xmlui/handle/123456789/7003
Title: Minimizing peak overshoot with reduced rule base on PID controller of switched reluctance motor drive [articol]
Authors: Muthuvinayagam, M.
Agalya, V.
Subjects: Multi layer Perceptrons
PID Controller
Switched Reluctance Motor Drive
Neuro Fuzzy Design Circuit
If-Then Rule
Peak Overshoot Time
Issue Date: 2019
Publisher: Timișoara : Editura Politehnica
Citation: Muthuvinayagam, M.; Agalya, V.: Minimizing peak overshoot with reduced rule base on PID controller of switched reluctance motor drive. Timişoara: Editura Politehnica, 2019.
Series/Report no.: Journal of Electrical Engineering;Vol 19 No 3
Abstract: Switched Reluctance Motor (SRM) is a machine where electromagnetic torque is designed and developed with variable changes in the reluctance process. Both stator and rotor has salient poles but only stator carries windings, hence a steady state error occurs in higher ratio on the Proportional-Integral-Derivative (PID) Controller. PID controller with variable gains on SRM consumes more peak overshoot time on the limited DC voltage supply saturation on specific devices. PID Fuzzy Controller for SRM shows the response time with very little bit variation over different load disturbances. But, the rise time on high speed electronics SRM drive fails to respond the circuit to fast input signals. All these limitations, planned to set up a SRM system design with PID Neuro Fuzzy Controller. To overcome high peak overshoot time on varying load disturbances, a framework called Neural Learning of PID Neuro Fuzzy based onCondensed Multi layer Perceptrons (NLPID Controller-NFCMP) is proposed in this paper. Initially, PID Neuro Fuzzy Controller uses the neural network learning techniques to tune the fuzzy membership function. The integration of neural network and fuzzy logic with the PID controller helps in handling multiple speed SRM drive machines. Here, multiple input points are taken and dynamic output system is generated with if-then rule fuzzification. Second, the Neuro fuzzy membership function with 49 rules in NLPID Controller-NFCMP framework is tuned to 12 rules using the Condensed Multi layer Perceptrons method. This proposed method is a feed forward artificial neural network which takes the previous if-then rule result to map the set using the activation function. Activation function in NLPID Controller-NFCMP framework condenses the rule based on the weighted input. As a result, the distributed load factor reduces the peak overshoot time and rise time. Experiment is carried out on the 8/6-pole SRM drive with varied load and speed and analyzed rising time, settling time and overshoot on load disturbances.
URI: https://dspace.upt.ro/xmlui/handle/123456789/7003
ISSN: 1582-4594
Appears in Collections:Articole științifice/Scientific articles

Files in This Item:
File Description SizeFormat 
BUPT_ART_Muthuvinayagam_f.pdf550.35 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.