ﻣﺪﻟﺴﺎزي داده راﻧﻪ ﻣﺪت زﻣﺎن ﺗﺪاوم وﻗﻔﻪ در ﺷﺒﮑﻪ ﺗﻮزﯾﻊ ﺑﺮق ﺑﺎ در ﻧﻈﺮ ﮔﺮﻓﺘﻦ ﻧﮕﻬﺪاري و ﺗﻌﻤﯿﺮات ﭘﯿﺸﮕﯿﺮاﻧﻪ، و ﺗﺤﻠﯿﻞ اﺛﺮ آن

Interruption Duration Modeling of Electrical Distribution Network Considering Preventive Maintenance and Analyzing Its Effects

ﯾﮑﯽ از ﻣﻌﯿﺎرﻫﺎي ﻣﻬﻢ در ﺳﻨﺠﺶ ﻋﻤﻠﮑﺮد ﺷﺮﮐﺖ ﻫﺎي ﺗﻮزﯾﻊ ﺑﺮق، ﻣﯿﺰان رﺿﺎﯾﺖ ﻣﺸﺘﺮﮐﯿﻦ اﺳﺖ. از دﯾﺪ اﮐﺜﺮ ﻣﺸﺘﺮﮐﯿﻦ ﻣﻬﻤﺘﺮﯾﻦ ﻣﻌﯿﺎر ﻋﻤﻠﮑﺮد اﯾﻦ ﺷﺮﮐﺖ ﻫﺎ، ﻣﺪت زﻣﺎن ﺗﺪاوم وﻗﻔﻪ ﻫﺎ اﺳﺖ. ﯾﮑﯽ از اﺑﺰاري ﻫﺎي ﺷﺮﮐﺖ ﻫﺎي ﺗﻮزﯾﻊ ﺑﺮق ﺑﺮاي ﮐﻨﺘﺮل و ﮐﺎﻫﺶ ﻣﺪت زﻣﺎن ﺗﺪاوم وﻗﻔﻪ ﻫﺎ ﻧﮕﻬﺪاري و ﺗﻌﻤﯿﺮات ﭘﯿﺸﮕﯿﺮاﻧﻪ اﺳﺖ. اﯾﻦ ﭘﮋوﻫﺶ ﺑﻪ ﻣﺪﻟﺴﺎزي ﻣﺪت زﻣﺎن ﺗﺪاوم وﻗﻔﻪ ﺑﺎ در ﻧﻈﺮ ﮔﺮﻓﺘﻦ ﻧﮕﻬﺪاري و ﺗﻌﻤﯿﺮات ﭘﯿﺸﮕﯿﺮاﻧﻪ ﻣﯽ ﭘﺮدازد. اﺑﺘﺪا داده ﻫﺎي وﻗﻔﻪ ﻫﺎي ﺳﯿﺴﺘﻢ از داﺷﺒﻮرد ﺑﻬﺮه ﺑﺮداري ﺷﺮﮐﺖ ﺗﻮزﯾﻊ ﻧﯿﺮوي ﺑﺮق ﺗﻬﺮان ﺑﺰرگ اﺳﺘﺨﺮاج ﮔﺸﺖ و ﺳﺎزﻣﺎﻧﺪﻫﯽ ﺷﺪ و ﺳﭙﺲ ﺑﺎ داده ﻫﺎي ﻫﻮاﺷﻨﺎﺳﯽ در ﺑﺎزه زﻣﺎﻧﯽ ﻣﻮرد ﺑﺮرﺳﯽ ﺗﻠﻔﯿﻖ ﮔﺮدﯾﺪ. در اداﻣﻪ ﭘﺲ از ﭘﺎﮐﺴﺎزي داده ﻫﺎ از روش ﻫﺎي رﮔﺮﺳﯿﻮن ﺧﻄﯽ و ﻏﯿﺮﺧﻄﯽ ﺑﺮاي ﻣﺪﻟﺴﺎزي ﻣﺪت زﻣﺎن ﺗﺪاوم وﻗﻔﻪ اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ. ﻧﻮآوري ﻫﺎ و ﻣﺸﺎرﮐﺖ ﻫﺎي اﯾﻦ ﭘﮋوﻫﺶ را ﻣﯽ ﺗﻮان: 1( ﻣﺪﻟﺴﺎزي داده راﻧﻪ ي زﻣﺎن ﺗﺪاوم وﻗﻔﻪ ﺑﺎ اﺳﺘﻔﺎده از اﻟﮕﻮرﯾﺘﻢ ﻫﺎي ﻣﺨﺘﻠﻒ ﺧﻄﯽ و ﻏﯿﺮﺧﻄﯽ، 2( در ﻧﻈﺮ ﮔﺮﻓﺘﻦ اﺛﺮ ﻣﺤﻞ اﺳﺘﻘﺮار ﺗﯿﻢ ﺗﻌﻤﯿﺮات و ﺗﺮاﻓﯿﮏ ﺷﻬﺮي ﺑﻪ ﺻﻮرت ﻣﺪت زﻣﺎن رﺳﯿﺪن ﺗﯿﻢ ﺗﻌﻤﯿﺮات ﺑﻪ ﻣﺤﻞ ﺣﺎدﺛﻪ، 3( در ﻧﻈﺮ ﮔﺮﻓﺘﻦ ﺗﺎرﯾﺨﭽﻪ اﻧﺠﺎم ﻋﻤﻠﯿﺎت ﻧﮕﻬﺪاري ﭘﯿﺸﮕﯿﺮاﻧﻪ در ﻣﺪﻟﺴﺎزي زﻣﺎن ﺗﺪاوم وﻗﻔﻪ و ﺗﺤﻠﯿﻞ ﺣﺴﺎﺳﯿﺖ زﻣﺎن ﺗﺪاوم وﻗﻔﻪ ﻧﺴﺒﺖ ﺑﻪ ﻋﻤﻠﯿﺎت ﻧﮕﻬﺪاري ﭘﯿﺸﮕﯿﺮاﻧﻪ، و 4( ﺗﺤﻠﯿﻞ ﺗﺎﺛﯿﺮ وﯾﮋﮔﯽ ﻫﺎي ورودي ﺑﺮ ﭘﯿﺶ ﺑﯿﻨﯽ زﻣﺎن ﺗﺪاوم وﻗﻔﻪ داﻧﺴﺖ.

Customer satisfaction is one of the important criteria in performance evaluation of electricity distribution companies. From the majority of customers’ point of view, interruption duration is the most important criterion. Electricity distribution companies control and reduce their interruption duration times using set of methods, including preventive maintenance. This paper develops an interruption duration model, considering preventive maintenance action. First, the system interruption data is extracted and organized from the operational dashboard in the Greater Tehran Electricity Distribution Company (GTEDC), and integrated with the meteorological data during the period under review. Then after cleaning the data, linear and nonlinear regression methods were used to model interruption duration. Innovations and contributions of this research can be considered as: 1) data-driven modeling of interruption duration using linear and nonlinear methods, 2) considering the effect of the repair crew locations and urban traffic, in terms of the repair crew travel time to the accident site, 3) considering the preventive maintenance operations history in modeling the interruption duration and sensitivity analysis of interruption duration to preventive maintenance, and 4) analyzing the influence of input features on interruption duration prediction