期刊名称：International Journal of Mechatronics, Electrical and Computer Technology
印刷版ISSN：2305-0543
出版年度：2014
卷号：4
期号：11
页码：602-609
出版社：Austrian E-Journals of Universal Scientific Organization
摘要：In this pap er we pro pos e an intelligen t met hod to detect fault diag nos is in the p hoto voltaic (PV) syst ems by applied the artificial neu ral n etwo rk (A NN). Firstly, the t emperatu re of the PV modu le is used to locat e the fau lt in the PV s ystem, an d us ually there is an obvio us t emperatu re difference between the fault and normal PV modu le. The current an d v oltag e of the maximum power po int tracking (MPPT) and the t emperatu re of the PV modu les are th e in p ut parameters of th e A NN, and the outp ut is the resu lt of the fault det ectio n. Th e simulation resu lt un der b oth normal and fault cond ition s sh ow t hat t he ou tput s of the A NN are almo st co nsist ent with t he expect ed v alue, and the prop osed fau lt diagnos is meth od can n ot on ly d etect and find the lo cation of the fault and determine the t ype of th e fau lt rap idly and accu rately.
