طراحي يك مدل بهينه‌سازي محدب براي خودترميمي يك شبكه توزيع هوشمند

Designing a Convex Optimization Model for Self-healing of a Smart Distribution Network

ﺧﻮدﺗﺮﻣﯿﻤﯽ ﺑﺮاي ﺗﺮﻣﯿﻢ ﺧﻮدﮐﺎر ﺳﯿﺴﺘﻢ ﺗﻮزﯾﻊ در ﻫﻨﮕﺎم ﺧﻄﺎ ﺑﻪ ﮐﺎر ﻣﯽرود. ﯾﮑﯽ از روشﻫﺎي ﮐﺎﻫﺶ ﻣﺸﺘﺮﮐﯿﻦ ﺑﺪون ﺑﺮق، ﺑﻬﺮه ﺑﺮداري ﻣﻨﻄﻘﻪ ﺧﻄﺎدار ﺑﻪﺻﻮرت ﺟﺰﯾﺮه اي اﺳﺖ اﻣﺎ زﻣﺎﻧﯽ ﺟﺰﯾﺮه ﺑﻪﺻﻮرت ﺑﻬﯿﻨﻪ ﺗﺸﮑﯿﻞ ﻣﯽﺷﻮد ﮐﻪ اﯾﻦ ﺟﺰﯾﺮه ﺑﻪﺻﻮرت آﻧﻼﯾﻦ ﺑﻌﺪ از ﺧﻄﺎ ﺷﮑﻞ ﮔﯿﺮد. در ﺻﻮرﺗﯽ ﮐﻪ ﺟﺰﯾﺮه ﻗﺒﻞ از ﺧﻄﺎ ﺗﻌﯿﯿﻦ ﺷﺪه ﺑﺎﺷﺪ، ﻣﺮز ﺑﻬﯿﻨﻪ ﺟﺰﯾﺮه ﺗﺸﮑﯿﻞ ﻧﻤﯽﺷﻮد و ﻣﻤﮑﻦ اﺳﺖ ﺗﻌﺪاد ﻣﺸﺘﺮﮐﯿﻦ ﺑﺪون ﺑﺮق اﻓﺰاﯾﺶ ﯾﺎﺑﺪ. در اﯾﻦ ﻣﻘﺎﻟﻪ ﺟﺰﯾﺮه ﺳﺎزي ﺑﻪﺻﻮرت آﻧﻼﯾﻦ ﭘﺲ از ﺧﻄﺎ اﻧﺠﺎم ﻣﯽﺷﻮد. ﯾﮑﯽ از ﻣﻬﻢﺗﺮﯾﻦ ﻣﺸﮑﻼت ﭘﺲ از وﻗﻮع ﯾﮏ رﺧﺪاد ﺧﻄﺎ، ﺣﻞ ﺳﺮﯾﻊ ﻣﺴﺌﻠﻪ اﺳﺖ. از آنﺟﺎﯾﯽ ﮐﻪ ﻣﺪلﻫﺎي ﻏﯿﺮﺧﻄﯽ، زﻣﺎن ﺣﻞ را ﺑﻪﺷﺪت زﯾﺎد ﻣﯽﮐﻨﻨﺪ؛ در روش ﭘﯿﺸﻨﻬﺎدي، ﻣﺴﺌﻠﻪ ﺧﻮدﺗﺮﻣﯿﻤﯽ از ﻣﺪل MINLP ﻧﺎﻣﺤﺪب ﺑﻪ ﯾﮏ ﻣﺪل ﻣﺤﺪب ﺗﺒﺪﯾﻞ ﺷﺪه اﺳﺖ. در ﻧﺘﯿﺠﻪ ﻣﺪل ﭘﯿﺸﻨﻬﺎدي ﺑﺎ زﻣﺎن ﺑﺴﯿﺎر ﮐﻤﯽ ﻣﯽﺗﻮاﻧﺪ ﺑﻪ ﺟﻮاب ﺑﺮﺳﺪ. ﻧﮑﺘﻪ دوم در ﻣﺴﺌﻠﻪ ﺑﻬﯿﻨﻪﺳﺎزي ﺧﻮدﺗﺮﻣﯿﻤﯽ رﺳﯿﺪن ﺑﻪ ﺟﻮاب ﺑﻬﯿﻨﻪ ﺳﺮاﺳﺮي اﺳﺖ. ﭼﻮن در ﻣﺴﺎﺋﻞ ﻣﺤﺪب ﺗﻨﻬﺎ ﯾﮏ راه ﺣﻞ ﺑﻬﯿﻨﻪ وﺟﻮد دارد؛ ﺑﻨﺎﺑﺮاﯾﻦ روش ﭘﯿﺸﻨﻬﺎدي رﺳﯿﺪن ﺑﻪ ﺟﻮاب ﺑﻬﯿﻨﻪ ﺳﺮاﺳﺮي را ﺗﻀﻤﯿﻦ ﻣﯽﮐﻨﺪ. ﻫﻤﭽﻨﯿﻦ ﻣﺪلﺳﺎزي ﺑﺎر و ﺗﻮﻟﯿﺪ در ﯾﮏ ﻓﻀﺎي اﺣﺘﻤﺎﻻﺗﯽ ﺻﻮرت ﻣﯽﮔﯿﺮد. ﭘﺲ از ﺗﻌﯿﯿﻦ رﯾﺰﺷﺒﮑﻪ ﺑﻬﯿﻨﻪ، ﺗﻮﻟﯿﺪ ﺑﻬﯿﻨﻪ ﻣﻨﺎﺑﻊ ﺗﻮﻟﯿﺪ ﭘﺮاﮐﻨﺪه و ﺑﺮداﺷﺖ ﺑﺎر ﺑﺮاي ﻣﺘﻌﺎدل ﮐﺮدن ﺗﻮان ﺳﯿﺴﺘﻢ اﻧﺠﺎم ﻣﯽﺷﻮد. ﺑﺮاي ﻧﺸﺎن دادن ﮐﺎراﯾﯽ روش ﭘﯿﺸﻨﻬﺎدي از ﺳﯿﺴﺘﻢ 33 ﺑﺎس IEEE اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ.

Self-healing is used to automatically repair the distribution system in the event of a fault. One way to reduce without service customers is to operate the faulty area as an island. However, the island is formed optimally when the island is formed online after the fault. If the island is determined before the fault, the optimal margin of the island will not be formed and the number of customers without electricity might increase. In this article, island building is carried out online after the fault. One of the most important problems after a fault occurs is the quick solution of the problem. Since nonlinear models greatly increase the solving time, in the proposed method, the problem is transformed from a non-convex MINLP model to a convex model. As a result, the proposed model can be answered in a very short time. The second point in the problem of self-healing optimization is to achieve the optimal global solution. The optimization problem is carried out by the convex optimization method. Because in convex problems there is only one optimal, the proposed method ensures the achievement of the optimal global solution. A variety of dispatchable and non-dispatchable distributed generation sources were used. Load and production modeling was also undertaken in a probabilistic space because considering the uncertainty, it can be said that the proposed model can be resistant to such events. After determining the optimal microgrid, the optimal production of distributed generation resources and load shedding was carried out to balance the system power. The IEEE 33-bus system was used to demonstrate the efficiency of the proposed method.