بهبود روش كنترل مستقيم گشتاور براي موتور القايي شش فاز با استفاده از روش هاي هوشمند

Improvement of the Direct Torque Control Method for a Sixphase Induction Motor Using Intelligent Methods

در اﯾﻦ ﭘﮋوﻫﺶ ﺑﻬﺒﻮد روش ﮐﻨﺘﺮل ﻣﺴﺘﻘﯿﻢ ﮔﺸﺘﺎور ﺑﺮاي ﻣﻮﺗﻮر اﻟﻘﺎﯾﯽ ﺷﺶ ﻓﺎز ﻣﻮرد ﺗﻮﺟﻪ ﻗﺮار ﮔﺮﻓﺘﻪ اﺳﺖ. اﯾﻦ روش ﯾﮑﯽ از ﮐﺎراﻣﺪ ﺗﺮﯾﻦ روش ﻫﺎي ﮐﻨﺘﺮﻟﯽ ﺑﺮاي ﻣﻮﺗﻮرﻫﺎي اﻟﮑﺘﺮﯾﮑﯽ اﺳﺖ ﮐﻪ ﻣﺤﺪودﯾﺖ ﻫﺎ و ﭘﯿﭽﯿﺪﮔﯽ ﻫﺎي ﮐﻤﺘﺮي ﻧﺴﺒﺖ ﺑﻪ روش ﮐﻨﺘﺮل ﺑﺮداري دارد، اﻣﺎ ﻣﺎﻧﻨﺪ ﺗﻤﺎم روش ﻫﺎي ﮐﻨﺘﺮﻟﯽ داراي ﻣﻌﺎﯾﺒﯽ ﻧﯿﺰ اﺳﺖ. اﺳﺘﻔﺎده از روش ﻫﺎﯾﯽ، ﻣﺎﻧﻨﺪ ﮐﻨﺘﺮل ﻫﺎي ﻏﯿﺮ ﺧﻄﯽ و اﻟﮕﻮرﯾﺘﻢ ﻫﺎي ﻫﻮﺷﻤﻨﺪ در ﺟﻬﺖ ﺑﻬﺒﻮد ﻧﺘﺎﯾﺞ روش ﮐﻨﺘﺮل ﻣﺴﺘﻘﯿﻢ ﮔﺸﺘﺎور ﺗﺎﺛﯿﺮ ﺑﺴﺰاﯾﯽ دارد. ﭘﺲ ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ اﯾﻨﮑﻪ ﺳﺎﺑﻘﻪ ﻃﺮاﺣﯽ روش ﮐﻨﺘﺮل ﻣﺴﺘﻘﯿﻢ ﮔﺸﺘﺎور ﺑﻪ دﻫﻪ 80 ﻣﯿﻼدي ﻣﯽ رﺳﺪ، ﺑﺮاي اﯾﻨﮑﻪ ﻧﺘﺎﯾﺞ ﺑﻪ دﺳﺖ اﻣﺪه ﺑﺎ ﻗﯿﺎس ﺑﻬﺘﺮي روﺑﺮو ﺷﻮد. اﺑﺘﺪا روش ﮐﻨﺘﺮل ﻣﺴﺘﻘﯿﻢ ﮔﺸﺘﺎور ﺑﺎ روش ﮐﻨﺘﺮل ﻣﻮد ﻟﻐﺰﺷﯽ ﺗﻄﺒﯿﻘﯽ ﺑﻪ اﺟﺮا در اﻣﺪه اﺳﺖ و در اداﻣﻪ ﺑﻪ ﺑﻬﺒﻮد ﻧﺘﺎﯾﺞ ﺑﺎ روش اﻧﻔﯿﺲ (ﻋﺼﺒﯽ _ ﻓﺎزي ﺗﻄﺒﯿﻘﯽ) ﭘﯿﺸﻨﻬﺎدي ﭘﺮداﺧﺘﻪ ﺷﺪه اﺳﺖ. روش ﻫﺎي ﻫﻮﺷﻤﻨﺪ از ﺟﻤﻠﻪ روش ﻫﺎﯾﯽ اﺳﺖ ﮐﻪ ﺑﻪ دﻟﯿﻞ ﺗﻮاﻧﺎﯾﯽ در ﺣﻞ ﻣﺴﺎﺋﻞ ﭘﯿﭽﯿﺪه، ﻏﯿﺮدﻗﯿﻖ و ﻏﯿﺮﺧﻄﯽ روز ﺑﻪ روز در ﺣﺎل اﻓﺰاﯾﺶ ﻫﺴﺘﻨﺪ. روش ﻋﺼﺒﯽ ﻓﺎزي ﺗﻄﺒﯿﻘﯽ از ﺟﻤﻠﻪ روش ﻫﺎي ﻫﻮﺷﻤﻨﺪ اﺳﺖ ﮐﻪ در اﯾﻨﺠﺎ ﺟﻬﺖ ﺑﻬﺒﻮد روش ﮐﻨﺘﺮل ﻣﺴﺘﻘﯿﻢ ﮔﺸﺘﺎور ﻣﻮرد اﺳﺘﻔﺎده ﻗﺮار ﮔﺮﻓﺘﻪ اﺳﺖ. ﺑﺎ اﺳﺘﻔﺎده از اﯾﻦ روش ﮐﻨﺘﺮﻟﯽ، ﺑﻪ ﺑﻬﺒﻮد ﭘﺎﺳﺦ ﮔﺸﺘﺎور و ﮐﺎﻫﺶ ﺧﻄﺎي ﺳﺮﻋﺖ ﭘﺮداﺧﺘﻪ ﺷﺪه اﺳﺖ. در روش ﮐﻨﺘﺮل ﻣﺴﺘﻘﯿﻢ ﮔﺸﺘﺎور ﺷﺎر ﮔﺸﺘﺎور ﻣﺤﺎﺳﺒﻪ و ﻣﯿﺰان ﺧﻄﺎي اﻧﻬﺎ ﺑﺎ ﻣﻘﺪار ﻣﺮﺟﻊ ﻣﻘﺎﯾﺴﻪ ﻣﯽ ﺷﻮد و ﺑﻪ ﮐﻨﺘﺮل ﮐﻨﻨﺪه اﻋﻤﺎل ﻣﯽ ﺷﻮد. ﺧﺮوﺟﯽ ﮐﻨﺘﺮل ﮐﻨﻨﺪه ﺑﻪ اﯾﻨﻮرﺗﺮ (وﻟﺘﺎژ) اﻋﻤﺎل ﻣﯽ ﺷﻮد ﺗﺎ اﯾﻨﻮرﺗﺮ ﺑﺮدار وﻟﺘﺎژ ﻣﻨﺎﺳﺐ ﺑﺮاي ﺗﻐﺬﯾﻪ ﻣﻮﺗﻮر را ﻓﺮاﻫﻢ ﮐﻨﺪ. ﻧﻮع ﻣﻮﺗﻮر اﻟﻘﺎﯾﯽ ﻣﻮرد ﻧﻈﺮ ﻧﺎﻣﺘﻘﺎرن ﻣﯽ ﺑﺎﺷﺪ. ﺳﯿﻢ ﭘﯿﭻ اﺳﺘﺎﺗﻮر ﯾﮏ ﻣﻮﺗﻮر اﻟﻘﺎﯾﯽ ﺷﺶ ﻓﺎز ﻧﺎﻣﺘﻘﺎرن از دو ﺳﯿﻢ ﭘﯿﭻ ﮐﻪ ﻧﺴﺒﺖ ﺑﻪ ﻫﻢ 30 درﺟﻪ اﺧﺘﻼف ﻓﺎز دارﻧﺪ ﺗﺸﮑﯿﻞ ﺷﺪه اﺳﺖ. ﭘﺲ ﻫﺪف از اﯾﻦ ﭘﮋوﻫﺶ اراﺋﻪ روش ﮐﻨﺘﺮل ﻫﻮﺷﻤﻨﺪ اﻧﻔﯿﺲ (ﻋﺼﺒﯽ _ ﻓﺎزي ﺗﻄﺒﯿﻘﯽ) ﺑﺮاي ﻣﻮﺗﻮر اﻟﻘﺎﯾﯽ ﺷﺶ ﻓﺎز ﻣﯽ ﺑﺎﺷﺪ واﻧﺘﺨﺎب ﺑﺮدار وﻟﺘﺎژ ﻣﻨﺎﺳﺐ ﺑﺮاي ﺑﻬﺒﻮد ﮔﺸﺘﺎور و ﺧﻄﺎي ﺳﺮﻋﺖ ﺗﻮﺳﻂ ﮐﻨﺘﺮل ﭘﯿﺸﻨﻬﺎدي ﻣﯽ ﺑﺎﺷﺪ. ﻧﺘﺎﯾﺞ ﺷﺒﯿﻪ ﺳﺎزي ﻧﻤﻮدارﻫﺎي ﺧﺮوﺟﯽ ﻣﺮﺑﻮط ﺑﻪ ﺷﺎر، ﺧﻄﺎي ﺳﺮﻋﺖ، ﺟﺮﯾﺎن و ﺳﺮﻋﺖ ﻣﻮﺗﻮر وﻫﻤﺠﻨﯿﻦ ﮔﺸﺘﺎور ﺧﺮوﺟﯽ را ﺑﺮاي ﻫﺮ دو ﮐﻨﺘﺮل ﻣﻮد ﻟﻐﺰﺷﯽ ﺗﻄﺒﯿﻘﯽ و ﻋﺼﺒﯽ - ﻓﺎزي ﺗﻄﺒﯿﻘﯽ ﻧﺸﺎن ﺧﻮاﻫﺪ داد؛ ﮐﻪ ﺑﻬﺒﻮد ﻧﺘﺎﯾﺞ در روش ﭘﯿﺸﻨﻬﺎدي ﻗﺎﺑﻞ ﻣﺸﺎﻫﺪه ﻣﯽ باشد.

This research has focused on improvement of the direct torque control method for six-phase induction motor. This method is one of the most efficient control methods for electric motors with fewer limitations and complexity than the vector control method, although it has some disadvantages like all other control methods. The use of methods such as nonlinear controls and intelligent algorithms has a significant effect on improving the results of the direct torque control method. Therefore, considering the fact that the direct torque control design dates back to the 80's, we have first compared the direct torque control method with the adaptive mode sliding control method so that the results could be better compared. Then we used the Adaptive Neuro-fuzzy method in order to improve the results. Intelligent methods are increasingly being used because they can solve complex, imprecise, and non-linear problems. The adaptive neuro-fuzzy method is an intelligent method that we have used in this research in order to improve the direct torque control method. Using this control method, we have improved the torque response and reduced the error rate. In the direct control method, the torque and flux are calculated and their error rates are compared with the reference value and applied to the controller. The output of the controller is applied to the inverter (voltage) so that the vector inverter can supply the appropriate voltage for the motor. The type of considered induction motor is asymmetrical. The stator coil of an asymmetric induction motor consists of two coils that have 30 degrees of phase difference. The purpose of this study is to provide an intelligent control method for a six-phase induction motor and select the appropriate voltage vector for improving the torque and speed rate using the proposed control. The simulation results will show the output graphs related to fluxes, speed error, the flow and engine speed, and the output torque for both adaptive sliding mode control and adaptive neuro-fuzzy control. The improved results in the proposed method are observable.