Abstract
Delo obravnava spremljanje in primerjavo obratovalnih
lastnosti sončnih elektrarn. Vsa odstopanja v delovanju lahko
spremljamo z dodatnimi meritvami na elektrarni, kot sta sončno
obsevanje in temperatura celice. Na podlagi polletnih meritev
parametrov delovanja je s pomočjo umetnega nevronskega omrežja v
programskem paketu Matlab pripravljen algoritem za izračun
napovedane moči sončne elektrarne v danem trenutku, s katerim lahko
ovrednotimo pravilno delovanje le-te. Omenjeni algoritem predstavlja
nadgradnjo sistema za spremljanje obratovanja sončne elektrarne.
Večja razlika med izmerjenimi in z algoritmom določenimi trenutnimi
izhodnimi močmi sončne elektrarne kaže na neustrezno delovanje
posameznih elementov sončne elektrarne in potrebo po podrobnejšem
preverjanju.
Keywords
algoritmi;napovedovanje;trenutna moč;sončne elektrarne;nevronsko omrežje;
Data
Other data
Secondary language: |
English |
Secondary title: |
Algoritem for predicting solar power plant output power with an artificial neural network |
Secondary abstract: |
This work deals with the comparison of operating propertis
of photovoltaic power plants. All derogations in the operation of
photovoltaic power plant can be monitored with additional
measurements of solar irradiation and temperature of photovoltaic
cells. Based on data acquired during six months operation of discussed
photovoltaic power plant an Artificial Neural Network (ANN) has been
built in order to predict output power of the power plant. The ANN
complements the already existing monitoring system. When the
difference between the ANN predicted and measured output power of
the photovoltaic power plant is too high, a detail check of the power
plant components is required. |
Secondary keywords: |
algorithm;prediction;out power;solar power plant;artificial neural network; |
URN: |
URN:SI:UM: |
Type (COBISS): |
Not categorized |
Pages: |
Str. [229]-238 |
ID: |
10869818 |