ﺑﺮرﺳﯽ ﻫﺎرﻣﻮﻧ ﯿﮏ ﻫﺎ در ﺷﺒﮑﻪ ﻫﺎي ﺗﻮز ﯾﻊ در ﺷﺮاﯾﻂ ﻧﺎﻣﺘﻘﺎرﻧﯽ ﺑﺎرﻫﺎي اﻟﮑﺘﺮﯾﮑﯽ

Harmonics Analysis in Distribution Networks under Asymmetric Conditions of Electrical Loads

ﺑﺎر ﻫﺎي ﻏﯿﺮﺧﻄﯽ د ر ﺷﺒﮑﻪ ﺗﻮزﯾﻊ ﻣﻮﺟﺐ ﻣﺸﮑﻼت ﻓﺮاواﻧﯽ ﺑﺮاي ﺷﺒﮑﻪﻫﺎي ﺗﻮزﯾﻊ ﻣﯽ ﺷﻮد. از ﺟﻤﻠﻪ اﯾﻦ ﻣﺸﮑﻼت ﻣﯽ ﺗﻮان ﺑﻪ ﺗﻠﻔﺎت، ﺗﺪاﺧﻞ در ﻋﻤﻠﮑﺮد ﺳﯿﺴﺘﻢﻫﺎي ﺣﻔﺎﻇﺘﯽ، ﮐﺎﻫﺶ ﻋﻤﺮ ﺗﺠﻬﯿﺰات و ﻋﻤﻠﮑﺮد ﻧﺎﻣﻄﻠﻮب ﻣﻮﺗﻮرﻫﺎ و ﺧﺎزن ﻫﺎ اﺷﺎره ﮐﺮد. در ﺳﺎﺧﺘﺎر ﺷﺒﮑﻪ ﻫﺎي ﺗﻮزﯾﻊ، ﺑﺎ رﺷﺪ اﺳﺘﻔﺎده از ﺗﺠﻬﯿﺰات اﻟﮑﺘﺮوﻧﯿﮑﯽ و و ﺟﻮد ﻣﺒﺪﻟﻬﺎي اﻟﮑﺘﺮوﻧﯿﮏ ﻗﺪرت، ﻣﻮﺟﺐ ﺗﺰرﯾﻖ ﻫﺎرﻣﻮﻧﯿﮏ و ﮐﺎﻫﺶ ﮐﯿﻔﯿﺖ ﺗﻮان در اﯾﻦ ﺷﺒﮑﻪ ﻫﺎ ﺷﺪه اﺳﺖ. اﺳﺘﻔﺎده از ﺗﺠﻬﯿﺰات ﻓﯿﻠﺘﺮ اﮐﺘﯿﻮ ﻣﯽﺗﻮاﻧﺪ ﻣﻮﺟﺐ ﮐﺎﻫﺶ ﻫﺎرﻣﻮﻧﯿﮏ و ﺑﻬﺒﻮد ﺷﺎﺧﺺﻫﺎي ﮐﯿﻔﯿﺖ ﺗﻮان در ﺷﺒﮑﻪ ﻫﺎي ﺗﻮزﯾﻊ ﮔﺮدد. در اﯾﻦ ﻣﻘﺎﻟﻪ، از اﻟﮕﻮرﯾﺘﻢ ﺑﻬﯿﻨﻪﺳﺎزي آﻣﻮزش و ﯾﺎدﮔﯿﺮي ﺑﻪ ﻣﻨﻈﻮر ﺑﻬﯿﻨﻪ ﺳﺎزي ﺿﺮاﯾﺐ ﮐﻨﺘﺮلﮐﻨﻨﺪه ﮐﻼﺳﯿﮏ ﺑﺎ ﻫﺪف ﮐﺎﻫﺶ ﻫﺎرﻣﻮﻧﯿﮏ اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ. ﻫﻢﭼﻨﯿﻦ از ﮐﻨﺘﺮلﮐﻨﻨﺪه ﻋﺼﺒﯽ ﭘﺮﺳﭙﺘﺮون ﭼﻨﺪ ﻻﯾﻪ ﺑﺎ اﻟﮕﻮرﯾﺘﻢ آﻣﻮزش ﻣﺒﺘﻨﯽ ﺑﺮ اﻟﮕﻮرﯾﺘﻢﻫﺎي ﻫﻮﺷﻤﻨﺪ ﺑﺮاي ﮐﺎﻫﺶ ﻫﺎرﻣﻮﻧﯿﮏ ﺷﺒﮑﻪ اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ. ﻣﻬﻢﺗﺮﯾﻦ وﯾﮋﮔﯽ ﺷﺒﮑﻪﻫﺎي ﻋﺼﺒﯽ، ﻏﯿﺮﺧﻄﯽ ﺑﻮدن و ﻗﺎﺑﻠﯿﺖ ﺣﺬف ﻧﻮﯾﺰ ﻣﯽﺑﺎﺷﺪ. ﺑﻪﻣﻨﻈﻮر ﺗﺎﯾﯿﺪ ﺑﺮ اﻧﺠﺎم ﮐﺎر از ﻣﺤﯿﻂ ﺑﺮﻧﺎﻣﻪ ﻧﻮﯾﺴﯽ و ﺷﺒﯿﻪ ﺳﺎزي ﻧﺮم اﻓﺰار ﻣﺘﻠﺐ اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ.

The existence of nonlinear loads in the distribution network causes many problems for distribution networks. These problems include losses, interference in the performance of protective systems, reduced lifetime of equipment and undesirable performance of motors and capacitors. In the structure of distribution networks, the growth of the use of electronic equipment and the presence of electronic converters have caused harmonic injection and reduced power quality in these networks. In this regard, the use of active filter equipment can reduce harmonics and improve power quality indices in distribution networks. In this paper, an optimization algorithm for teaching and learning is proposed to optimize the classical controller coefficients with the aim of reducing harmonics. Also, Multilayer Perceptron Neural Controller has been used to train algorithm based on intelligent algorithms to reduce network harmonics. The most important features of neural networks are nonlinearity and noise cancellation. In order to perform the simulation of software, the programming environment and simulation of MATLAB software is used.