Short-term prediction for distributed photovoltaic power based on improved similar time period
A short-term prediction method for distributed PV power based on an improved selection of similar time periods (ISTP) is proposed, to address the problem of low output power prediction accuracy due to a large number of influencing factors and the large difference in the degree of influence of variou...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2023-02-01
|
Series: | Frontiers in Energy Research |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fenrg.2023.1149505/full |
_version_ | 1797897079108927488 |
---|---|
author | Zhihan Wu Yi Zhang Bijie Liu Minghui Zhang |
author_facet | Zhihan Wu Yi Zhang Bijie Liu Minghui Zhang |
author_sort | Zhihan Wu |
collection | DOAJ |
description | A short-term prediction method for distributed PV power based on an improved selection of similar time periods (ISTP) is proposed, to address the problem of low output power prediction accuracy due to a large number of influencing factors and the large difference in the degree of influence of various factors. First, the simple correlation coefficient (SCC) based on path analysis is used to screen the main influencing factors with stronger correlation with PV output power, and these factors are classified into three categories. Second, correlations of the three dimensions are calculated, respectively: (i) TOPSIS (with weights optimized by the SCC) determines meteorological correlation, (ii) linear weighting (based on the fuzzy ranking) obtains time correlation, and (iii) load correlation is quantified with existing current parameters. Third, the combined impact correlation (CIC) is obtained by weighting the three correlations above to establish criteria for the selection of similar periods, and a short-term PV power prediction model is established. Finally, experimental results based on real data of Australian Yulara Solar System PV plant demonstrate that errors of proposed ISTP method are respectively improved by 47.06% and 46.09% compared with the traditional ELMAN and similar day method. |
first_indexed | 2024-04-10T07:52:01Z |
format | Article |
id | doaj.art-6291bb40b71b4eab9777ed0e40f48cbc |
institution | Directory Open Access Journal |
issn | 2296-598X |
language | English |
last_indexed | 2024-04-10T07:52:01Z |
publishDate | 2023-02-01 |
publisher | Frontiers Media S.A. |
record_format | Article |
series | Frontiers in Energy Research |
spelling | doaj.art-6291bb40b71b4eab9777ed0e40f48cbc2023-02-23T09:25:22ZengFrontiers Media S.A.Frontiers in Energy Research2296-598X2023-02-011110.3389/fenrg.2023.11495051149505Short-term prediction for distributed photovoltaic power based on improved similar time periodZhihan WuYi ZhangBijie LiuMinghui ZhangA short-term prediction method for distributed PV power based on an improved selection of similar time periods (ISTP) is proposed, to address the problem of low output power prediction accuracy due to a large number of influencing factors and the large difference in the degree of influence of various factors. First, the simple correlation coefficient (SCC) based on path analysis is used to screen the main influencing factors with stronger correlation with PV output power, and these factors are classified into three categories. Second, correlations of the three dimensions are calculated, respectively: (i) TOPSIS (with weights optimized by the SCC) determines meteorological correlation, (ii) linear weighting (based on the fuzzy ranking) obtains time correlation, and (iii) load correlation is quantified with existing current parameters. Third, the combined impact correlation (CIC) is obtained by weighting the three correlations above to establish criteria for the selection of similar periods, and a short-term PV power prediction model is established. Finally, experimental results based on real data of Australian Yulara Solar System PV plant demonstrate that errors of proposed ISTP method are respectively improved by 47.06% and 46.09% compared with the traditional ELMAN and similar day method.https://www.frontiersin.org/articles/10.3389/fenrg.2023.1149505/fulldistributed photovoltaic power plantpath analysisTOPSISPV power predictionimproved similar period |
spellingShingle | Zhihan Wu Yi Zhang Bijie Liu Minghui Zhang Short-term prediction for distributed photovoltaic power based on improved similar time period Frontiers in Energy Research distributed photovoltaic power plant path analysis TOPSIS PV power prediction improved similar period |
title | Short-term prediction for distributed photovoltaic power based on improved similar time period |
title_full | Short-term prediction for distributed photovoltaic power based on improved similar time period |
title_fullStr | Short-term prediction for distributed photovoltaic power based on improved similar time period |
title_full_unstemmed | Short-term prediction for distributed photovoltaic power based on improved similar time period |
title_short | Short-term prediction for distributed photovoltaic power based on improved similar time period |
title_sort | short term prediction for distributed photovoltaic power based on improved similar time period |
topic | distributed photovoltaic power plant path analysis TOPSIS PV power prediction improved similar period |
url | https://www.frontiersin.org/articles/10.3389/fenrg.2023.1149505/full |
work_keys_str_mv | AT zhihanwu shorttermpredictionfordistributedphotovoltaicpowerbasedonimprovedsimilartimeperiod AT yizhang shorttermpredictionfordistributedphotovoltaicpowerbasedonimprovedsimilartimeperiod AT bijieliu shorttermpredictionfordistributedphotovoltaicpowerbasedonimprovedsimilartimeperiod AT minghuizhang shorttermpredictionfordistributedphotovoltaicpowerbasedonimprovedsimilartimeperiod |